�

�

����������� ��� ��� �

� ����������������� �����

� ������

� ����� �������������

� ��������� ��� ����

�

National Forest and Tree Assessment and Inventory TCP/LEB/2903

Final Report

September 2005

Compiled by: Mr. Soren Dalsgaard International Forestry Consultant TCP/LEB/2903

2

I. TABLE OF CONTENTS I. TABLE OF CONTENTS ...............................................................................................................2 II LIST OF ACRONYMS AND ABREVIATIONS........................................................................5 III EXECUTIVE SUMMARY ..........................................................................................................6 IV FOREWORD ...............................................................................................................................12 1. Introduction ..................................................................................................................................13

2. Background and justification......................................................................................................15 2.1 General background ..............................................................................................................15 2.2 Existing datasets .....................................................................................................................18

2.2.1 The 1962-1965 1:50.000 Forest Type Map of Lebanon. ..................................................18 2.2.2 The 2000-2002 mapping of the Landcover / Landuse (LCLU) under TCP/LEB/2801.....19

2.3 Justification.............................................................................................................................22

3. Objectives of the Study ................................................................................................................23

4. Resources for the implementation of the NFA ..........................................................................24 4.1 Institutions ..............................................................................................................................24 4.2 Infrastructure .........................................................................................................................25 4.3 Staff .........................................................................................................................................25

4.3.1 MOA Staff..........................................................................................................................25 4.3.2 External Staff.....................................................................................................................26

4.4. Equipment..............................................................................................................................27 4.4.1 Field equipment.................................................................................................................27 4.4.2 Hardware - Forest Information System ............................................................................27 4.4.3 Software - Forest Information System ..............................................................................28 4.4.4 Digital Maps and satellite images ....................................................................................28

5. Project Management....................................................................................................................30 5.1 National Project Coordinator ...............................................................................................30 5.2 National Consultants and Team Leaders.............................................................................31 5.3 International Consultants and FAO Support......................................................................32 5.4 Project Steering Comittee .....................................................................................................34

6. Preparations .................................................................................................................................35 6.1 Recruitment of national and international consultants......................................................35 6.2 Interpretation of satellite images and map construction....................................................35 6.3 Procurement of forest inventory and information system equipment ..............................36 6.4 Memorandum of Understanding between DRDNR and DSC ...........................................37 6.5 Field manuals and field forms...............................................................................................37

7. Methodology .................................................................................................................................38 7.1 Sampling design, classification system and variables .........................................................38

7.1.1 Workshop on national forest inventory methodology .......................................................38 7.1.2 Workshop on information needs and variables.................................................................39 7.1.3 Standardized and harmonized vegetation classification system .......................................39 7.1.4 Sampling design ................................................................................................................41 7.1.5 Variables ...........................................................................................................................43

3

7.2 Training...................................................................................................................................43 7.2.1 Workshops .........................................................................................................................43 7.2.2 Study Tour .........................................................................................................................44 7.2.3 Training of Engineers and Team Leaders.........................................................................44 7.2.4 Training of Field Teams....................................................................................................45

7.3 Production of a derived map of forest and TOF resources of Lebanon ...........................45 7.3.1 Methodology for mapping .................................................................................................46 7.3.2 Output of Mapping ............................................................................................................47

8. Fieldwork and coordination ........................................................................................................49 8.1 Distribution of tracts to Mohafazas......................................................................................49 8.2 Fieldwork timing and supervision ........................................................................................51 8.3 Field work experiences ..........................................................................................................51 8.4 Coordination of activities ......................................................................................................54

9. Database design, dataentry, processing and analysis ...............................................................55 9.1 Database Design and data structure.....................................................................................55 9.2 Data entry ...............................................................................................................................55 9.3 Data processing ......................................................................................................................56 9.4 Data analysis ...........................................................................................................................56

10 Results ..........................................................................................................................................57 10.1 Area .......................................................................................................................................57

10.1.1 Area by forest types and Land Use Classes ....................................................................57 10.1.2 Area by Ecological Zone.................................................................................................63 10.1.3 Proportion of forest area by management system ..........................................................63 10.1.4 Proportion of forest area for each protection level ........................................................64 10.1.5 Proportion of forest area by ownership ..........................................................................65 10.1.6 Precision of area estimates .............................................................................................65

10.2 Volume ..................................................................................................................................66 10.2.1 Gross Volume of Global Classes ....................................................................................66 10.2.2 Precision of volume estimates.........................................................................................67

10.3 Biomass .................................................................................................................................68 10.3.1 Biomass of forest species ................................................................................................68 10.3.2 Biomass of TOF species ..................................................................................................69 10.3.3 Above ground woody biomass in Forest and TOF by ecological zone...........................70 10.3.4 Precision of biomass estimates .......................................................................................70

10.4 Carbon Stock ........................................................................................................................71 10.4.1 Carbon Stock by Global Classes.....................................................................................71 10.4.2 Precision of Carbon storage estimates ...........................................................................71

10.5 Dbh. Distribution .................................................................................................................72 10.6 Biodiversity ...........................................................................................................................73

10.6.1 List and frequency of tree species in forest by forest type ..............................................73 10.6.2 List and frequency of tree species outside forest (TOF) .................................................76 10.6.3 Forest by stand structure option .....................................................................................80 10.6.4 Forest by human induced disturbance level ...................................................................81

4

10.7 Use of resources ....................................................................................................................82 10.7.1 Wood and NWFP provided by the forest ........................................................................83 10.7.2 Wood and NWFP provided by the TOF ..........................................................................83 10.7.3 Social, economic and environmental services provided by the forest ............................85 10.7.4 Social, economic and environmental services provided by TOF ....................................86 10.7.5 Users of forests and tree resources.................................................................................87 10.7.6 Reliability of estimates of forest use attributes ...............................................................88

11. Forest Fires .................................................................................................................................89

12. Local Population ........................................................................................................................90 12.1 Trends ...................................................................................................................................90 12.2 Activities................................................................................................................................90

13. Accessibility to tracts and plots.................................................................................................93 13.1 Accessibility to hospitals and schools .................................................................................93 13.2 Accessibility to Settlements and markets...........................................................................94 13.3 Accessibility to all weather roads and seasonal roads ......................................................95

14. Comparison of NFA results with existing information...........................................................97

15. Capacity building .......................................................................................................................99 15.1 Capacity to undertake fieldwork ........................................................................................99 15.2 Capacity to register and enter data ....................................................................................99 15.3 Capacity to validate and merge data................................................................................100 15.4 Capacity to process and analyse data...............................................................................100 15.5 Capacity to produce maps .................................................................................................101 15.6 Capacity to organize and coordinate an NFA .................................................................101

16. Time and cost analysis .............................................................................................................102

17. Conclusions ...............................................................................................................................103

18. Recommendations ....................................................................................................................105 18.1 Future Monitoring of Resources.......................................................................................105 18.2 Additional Information Needs ..........................................................................................107 18.3 Actions for Resources Development .................................................................................107 18.4 International Reporting.....................................................................................................108 18.4 International Reporting.....................................................................................................108

19. Bibliography .............................................................................................................................109 Appendix 1: Statistical functions used in data analysis..............................................................111

5

II. LIST OF ACRONYMS AND ABREVIATIONS

AFDC Association for Forest Development and Conservation CI Climatic Index COFO Committee on Forestry DFM Derived Forest Map DRDNR Directorate of Rural Development and Natural Resources DSC Directorate of Studies and Coordination FAO Food and Agriculture Organization of the United Nations FD Forest Department FORM Forest Resources Development Service FRA Forest Resources Assessment GIS Geographic Information System GPS Global Positioning System IC International Consultant LCCS Land Cover Classification System LCLU Land Cover / Land Use LCLUP Land Cover / Land Use Project LNU Lebanese National University MOA Ministry of Agriculture MOE Ministry of Environment MoU Memorandum of Understanding NC-FI National Consultant Forest Inventory NC-ID National Consultant Institutional Development NC-RSM National Consultant Remote Sensing and Mapping NCSR National Council of Scientific Research NFA National Forest Assessment NFAP National Forest Action Plan NGO Non Governmental Organization NPC National Project Coordinator NRRDD Natural Resources and Rural Development Department NWFPS Non-wood Forest Products and Services OL Other Land OLWT Other Land With Trees OWL Other Wooded Land RDS Regional Development Services SC Steering Committee TCDC Technical Cooperation Between Developing Countries TCP Technical Cooperation Programme TOF Trees Outside the Forest TOR Terms of Reference USJ Université St. Joseph

6

III EXECUTIVE SUMMARY Background and Justification: The first National Forest and Tree Assessment based on systematic sampling in Lebanon was implemented by the Directorate of Rural Development and Natural Resources (DRDNR) of the Ministry of Agriculture as a project (TCP/LEB/2903) during 2003 - 2005 under the Technical Cooperation Programme of the Food and Agriculture Organization of the United Nations (FAO). Timely, easily available and accurate information on Forest and Trees Outside the Forest (TOF) resources and their utilization is a precondition for sustainable management of these resources based on an environmentally, socially and economically balanced forest policy. Prior to TCP/LEB/2903 the extent and quality of information on the forest and TOF resource of Lebanon was perceived inadequate to monitor and counteract resources depletion and environmental degradation and to support national policy decisions and the work on developing a National Forest Action Plan. The Government of Lebanon therefore requested the technical assistance of FAO under the Technical Cooperation Programme to assist the Ministry of Agriculture to develop capabilities to survey the country’s resources to monitor trends and changes, to describe their nature and localities and to advise on protective measures where required for resources at risk. The National Forest and Tree Assessment (NFA) was undertaken by the DRDNR and was implemented in the field by the Rural Development Service (RDS), in collaboration with the Forestry Department of FAO. The main objective of the project was to reinforce the capacity of DRDNR for collecting, compiling, analyzing and disseminating reliable and up–to-date information on the Forest and TOF resources of Lebanon through training of the national staff in forest and tree inventory techniques. Project Phases: According to the Project Document TCP/LEB/2903 was characterized by 3 well defined phases:

Phase I July 2003 - November 2003: Preparation, Training and Mapping; Phase II November 2003 - July 2004: Field Survey; Phase III July 2004 - September 2005: Data Processing and Reporting.

Phase I: Included all preliminary arrangements and preparatory work related to acquisition of equipment, recruitment of staff, production of field manuals, field forms and working maps, obtaining access to geographic data, training sessions for staff, adjustment of methodology and harmonizing vegetation classification system and variables according to national requirements. Phase II: The fieldwork was conducted during the second project phase. The field measurements and interviews were conducted by 5 field teams that were based in the Rural Development Services.

A systematic grid of 222 permanent sample sites (0.04% sampling intensity) was established in accordance with the FAO approach to supporting NFA. On the systematic sampling grid, data on the Forest and TOF resources and information concerning the use of wood and non-wood-forest products and services (NWFPS) was collected, in accordance with the standard FAO-specifications supplemented by the additional national requirements for statistics on the state of the resources. After double checking the data was entered into the database (Forest Information System) which was also developed during Phase II. Phase III: During the third project phase the collected data was processed, analysed and interpreted. A Derived Map of the Forest and TOF resources of Lebanon was completed after reinterpreting the existing spatial database from the Land Cover Land Use Map (1998) supported by the best available georeferenced information on the forest and TOF resources of Lebanon and a period of intensive field validation. The third Project Phase also included presenting the project findings to the public.

7

Project Findings Two area datasets were produced by the project. One based on the results from the field survey (NFA) and one based on the derived map of forest and TOF resources of Lebanon (DFM). The area findings from the project at the global class level (Level 1) are presented below:

Country area by FRA Global Classes from DFM and NFA

0

10

20

30

40

50

60

70

80

90

100

DFM NFA

Per

cent

Inland Water

Other Land

Other Wooded Land

Forest

FRA Global class NFA

Area (ha) DFM Area (ha)

Forest 139,376 134,372 Other Wooded Land (OWL) 108,378 120,574 Other Land (OL) 797,152 766,289 Inland Water 294 1745

The datasets show similar percentages at the global class level. The NFA estimates the forest area of Lebanon at 13.3 % of the total area of the country, while the DFM determines the forest area at 13.1 %.

The NFA estimates the OWL area of Lebanon at 10.4 % of the total area of the country, while the DFM determines the OWL area at 11.8 %. Forest area by forest type For the main forest types (Level 2) of the vegetation classification system the findings are as follows:

Percentage of forest area by forest type (Level 2) according to DFM and NFA. Forest Type Level 2 DFM (%) NFA (%) Coniferous Forest 34.2 32.2 Broadleaved Forest 53.1 56.6 Mixed Forest 12.7 11.2 Total 100 100

The findings from the DFM and the NFA for the main forest area also compare well. Both indicate that slightly more than half the forest area of Lebanon is broadleaved forest; a third of the forest area is coniferous forest while the remaining 11-13 % is mixed forest.

8

Forest area by Level 3 and 4 forest type according to DFM and NFA.

0

20000

40000

60000

80000

100000

120000

140000

160000

DFM NFA

hect

ares

Mixed forest

Broadleaved decidous

Broadleaved mixed

Broadleaved evergreen

Fir

Cypressus

Cedars

Mixed coniferous

Pinus pinea

Juniper

Other pines

When comparing the findings at level 3 and 4 the main differences between the DFM and the NFA findings are related to the categories Broadleaved mixed and Broadleaved evergreen and the fact that the NFA did not capture any fir areas in the sampling grid. Rare occurrences may not be captured by a sampling grid of 0.04% sampling intensity.

OWL area by OWL type from DFM and NFA

0

20000

40000

60000

80000

100000

120000

140000

DFM NFA

hect

ares

OWL without trees

OWL with trees

Coniferous shrubs

Mixed shrublands

Grassland with trees

Broadleaved shrub

The total OWL areas are similar in the two data sets. During the production of the Derived Forest and TOF map it became apparent that it was not possible to distinguish between the various OWL classes employed by the NFA, hence the need to employ a different subdivision of OWL for the mapping of OWL (OWL with trees and OWL without trees).

9

Volume and biomass The NFA indicates an average above ground standing volume of wood at app. 35.6 m3/ha for forest, 5.1 m3/ha for OWL and 3.7 m3/ha for Other Land. The NFA estimates of national totals for biomass and carbon stock at the global class levels are: Above ground

biomass (million tons)

Below ground biomass

(million tons)

Total biomass (million tons)

Carbon from Above ground

biomass (million tons)

Carbon from Below ground

biomass (million tons)

Carbon from Total biomass (million tons)

Forest 2.793 0.797 3.59 1.397 0.399 1.79 OWL 0.325 0.082 0.407 0.163 0.041 0.20 OL 2.412 0.649 3.061 1.206 0.325 1.53 Forest and OWL characteristics according to NFA Species distribution In broadleaved forests Quercus species, are the most common (Q. calliprinos, Q. cerris and Q. infectoria). In coniferous forests Pinus species are the most common (P. brutia and P. pinea) Trees outside forests are mainly olives and orchard species (Olea europea, Citrus spp., Malus spp. and Prunus spp.) Stand structure Almost all broadleaved forest has a structure with a single canopy layer, while in coniferous forest, (especially in pruned P. pinea stands) a two layered stand structure can develop. Disturbances Almost all forest in Lebanon is classified as disturbed (85%) – which in part can be due to the long history of settlement and forest utilization in Lebanon. The NFA indicates that only a small fraction of the Forest and TOF areas are formally protected. Ownership. The NFA indicated that a large percentage of the Forest (60%) and TOF areas (80%) are under private ownership. As this is based on a 0.04% sample it should only be taken as an indicator of a high degree of private ownership. Products Fuelwood is the most common product that is extracted from forests and OWL. Plant food is the second most common in both Forest and OWL. Services Forest and OWL provide numerous products and services among which soil and water conservation and grazing rank highest. Soil and water conservation is the most common service provided by the forest in Lebanon, grazing is ranked second. In OWL the trend is opposite with grazing being the most common service. Forest fires 14.1 % of the forest area surveyed during the fieldwork showed evidence of fires. Forest fires are almost exclusively limited to coniferous stands and almost always develop from fires started by humans. No evidence of forest fires was registered in broadleaved forests.

10

Conclusion and recommendations TCP/LEB/2903 represents the first national experience with National Forest and TOF Resources Assessment based on systematic field sampling in Lebanon. The current NFA project should be seen as a pilot study. It provides a first snapshot of the state of the forest resources in Lebanon based on systematic ground measurements. The experience from the current NFA shows that the approach promoted by FAO is a cost efficient way of building national capacity to generate and analyse data on the forest and TOF resources – in many cases the data relates to aspects of Forests and TOF which have not earlier been investigated. The similarity of the area-datasets of the NFA and the DFM at the first and second level of vegetation classification indicate that the NFA area estimates at these levels are representative. By investing about 17.000 man hours in the NFA-related work, a primary set of data on Forests and TOF has been generated, the capacity of DRDNR has been built in all aspects related to forest inventory and mapping and a permanent systematic sampling system has been set up which allows for future monitoring. The data set is broader in scope than earlier studies of the Lebanese Forest and TOF resource and it is based on a systematic sampling grid and a methodology that allows for future repetitions and expansions of scope. The intention of DRDNR is to repeat the National Forest and TOF assessment as a periodic inventory at regular intervals within the institutional framework of the Directorate. With repetitions, the NFA will capture information related to the change in the extent, state and uses of the forest and TOF resources and it will be possible to monitor the changes. Additional parameters can be included for future measurements if deemed relevant. The value will therefore increase with every re-measurement. The permanent sampling grid could offer a platform for collaboration with other institutions for expanding the scope and increasing the level of detail of the NFA. Among areas that could be investigated further are the following: � Development of more appropriate methods of measuring growing stock in orchard trees. � Field measurement of vegetation classes that were not captured by the field inventory (e.g. fir). � Forest and TOF maps (i.e. detailed mapping based on aerial photos and maps for management purposes). � Studies for elaboration of volume in dense shrubs and coppices � Additional studies of Products and Services from Forest and TOF for management purposes. � Additional species and location specific studies in the relationship between age and growing stock The current NFA estimates the forest area of Lebanon at around 13 % of the total area of the country and the OWL area of Lebanon at 10-12 %. Former reports and estimates showed figures for forest areas between 5 and 7%. The findings of TCP/LEB/2903 are therefore in sharp contrast to the previous generally quoted datasets for reporting on the forest cover of Lebanon. As both the NFA and the DFM arrive at very similar figures at the Global Class Level it is likely that the forest and TOF resources of Lebanon reported for FRA 2000 were underestimated. The difference between former estimations and the result of the NFA could be attributed to the following (BEYDOUN & STEPHAN 2005): • The fact that a dataset from the 1960s was used for reporting in 2000 - Rapid urbanization with a

considerable rural migration caused abandoning of agricultural lands and pastoral activities followed by invasion of many areas by forest during the second part of the 20th century.

• Differences in definitions for “forest” and “other wooded lands” for Lebanon in earlier estimates and possibly also unclear definitions of OWL.

• Reforestation activities and relative protection of many coniferous stands from cutting and felling during the past decades.

11

The NFA should be seen as a starting point for a tradition of conducting systematic inventories and using the gathered information on the forest and TOF resources in the development of appropriate national policies and plans and their subsequent implementation, i.e. a continual process as illustrated below.

The ultimate value of the NFA will depend on how useful the captured data is. Constructive follow up on the NFA is important in order to ensure that the data captured is relevant for meeting national needs. Developing the correct policies depends on timely and relevant information. The NFA is an important step for providing such information and will continue to evolve as the knowledge of the Forest and Tree Resources continue to evolve. The findings of the current NFA can be instrumental in a debate about the extent of knowledge about the Forest and TOF Resources and the importance of access to timely and relevant data on the Forest and TOF resource and function as a catalyst for more research. The NFA should therefore not be seen as the end product in itself – but the start of a process.

12

IV FOREWORD Lebanon is the fifth country in the world to implement a National Forest Assessment (NFA) according to the guidelines developed by FAO Forest Resources Development Service (FORM). Through the activities of project (TCP/LEB/2903) the Ministry of Agriculture, Directorate of Rural Development and Natural Resources has established a nationwide network of permanent sample sites where on-the-ground information of the Forest and TOF resources of Lebanon has been measured. This provides timely and relevant information concerning the state and use of the forest and TOF and it allows monitoring changes through future re-measurements. Through the NFA, a wide range of field data related to forests and trees has been registered throughout Lebanon. This information is valuable not only to the managers of the forests and TOF, but it is also an important tool for planning and policy purposes and as a stepping stone for further research. Many potential areas for fruitful future collaboration with other Institutions can therefore be foreseen to develop the NFA even further. Through the NFA activities a substantial capacity building has been undertaken at MOA / DRDNR. 10 Engineers and 57 Forest guards were trained in forest assessment techniques. The fieldwork was undertaken by 5 Main Crews + one additional crew. Through the continual on-the-job-training of the fieldwork phase the skills were perfected. Through the NFA, the capacity of the Ministry of Agriculture, Directorate of Rural Development and Natural Resources has been boosted in the areas of mapping as well as data processing and analysis. The NFA gives data on the Lebanese Forest and TOF resource and it is a useful tool for the development of a National Strategy and Policy for the Forest and TOF resources of Lebanon. Through the combined skills and dedication of all staff involved a foundation has now been laid for a tradition of systematic sampling of the forest and tree resources of Lebanon. In the future National Forest Assessment will be part of the ministry activities, as it should be repeated periodically in order to get more data needed for a national forest and TOF strategy. Thanks to all the Institutions for collaboration, especially to the Lebanese Army which through the Geographic Affairs Directorate have been of great help in providing maps for the project.

Mr. Ghattas Akl National Project Coordinator

Director, DRDNR

13

1. Introduction. The Food and Agriculture Organization of the United Nations (FAO), at the request of member nations and the world community, regularly monitors the world’s forests through the FRA Programme. National forest definitions and the survey methods used in data collection vary between countries. The differences in definitions and measurement methodologies have resulted to problems of comparability of national data and reliability of aggregated results. Likewise, many countries still lack reliable primary technical information at the country level. To help address the issue the FRA Programme, following the 2001 meeting of the Committee on Forestry (COFO), launched a new approach to support national forest assessment (NFA) through in-country capacity building activities. The long-term objective of the FAO-NFA approach is “to contribute to the sustainable management of forests and trees outside forest by providing decision makers and stakeholders with the best possible, most relevant and cost effective information for their purposes at local, national and international levels.” The FAO-NFA approach to NFA was first pilot-implemented in Costa Rica and later replicated in Guatemala, Cameroon, the Philippines, and then Lebanon through TCP/LEB/2903 (CONSOLACION, C.P., 2004, FAO 2001). TCP/LEB/2903 National Forest and Tree Inventory and Assessment of Lebanon became active in July 2003. The main objective of the project was to reinforce the capacity of the Directorate of Rural Development and Natural Resources (DRDNR) for collecting, compiling, analyzing and disseminating reliable and up–to-date information on the forest and trees outside the forest (TOF) resources of Lebanon through training of the national staff in forest and tree inventory techniques (MOA/FAO 2003). During the project, a systematic grid of permanent sample plots was established in accordance with the forest inventory component of the FAO-NFA approach. Tracts for sampling (1 km x 1km) were established at every 4 minutes longitude and latitude (every 6-7 km) across Lebanon. On the systematic grid of tracts, data on the forest and trees outside forest (TOF) resources and information concerning the use of wood and non-wood-forest products and services (NWFPS) was measured, registered and analyzed in accordance with the specifications of the FAO-FORM support to National Forest Assessments and the additional national requirements for accurate statistics on the state of the resources. Information on the supply and demand trend of wood and NWFPS, and the uses of such resources was collected through interviews with the local forest users who extract forest products from the sites measured and/or who have information about the forest products extracted. The interviews provide valuable information which are difficult, if not impossible, to acquire through direct observations of the forest resources (CONSOLACION, C.P., 2004). TCP/LEB/2903 is characterized by 3 well defined phases:

Phase I July 2003-November 2003: Preparation, Training and Mapping; Phase II November 2003-July 2004: Field Survey; Phase III July 2004-September 2005: Data Processing and Reporting.

14

TCP/LEB/2903 provides the first experiences with systematic field sampling of the forest and TOF resources of Lebanon. The current report is the Final Report TCP/LEB/2903 Forest and Tree Inventory and Assessment of Lebanon, September 2005. The report summarizes the work undertaken during TCP/LEB/2903 and the findings of the project.

15

2. Background and justification 2.1 General background The Republic of Lebanon is situated on the Eastern shores of the Mediterranean Sea, approximately between 35.2 - 36.6 E and 33.1 - 34.7 N. Lebanon covers a total of 10452 km2. The topography is characterized by the Mount Lebanon and the Anti-Lebanon mountain chains that run parallel to the coast and are separated by the Beqaa Valley. 1 Please refer to figure 3 for a topographic map of Lebanon. The ecological conditions of Lebanon are determined largely by topography and vary with altitude and exposition. The climatic conditions vary from Mediterranean climate along the coast and the mid altitudes of the mountain ranges, via sub alpine or mountain Mediterranean climate on the highest slopes to arid / sub-desertic in the northern plains. Most rainfall falls between November and March. The mean annual rainfall on the coast ranges between 700 – 1000 mm. The central part of the Mount Lebanon coastal range receives up to 1600 mm annually. In the Beqaa Valley the rainfall ranges from 200 mm in the north east in the rain shadow of Mt. Lebanon mountainrange to 800 mm in the south. The Anti Lebanon receives between 600 mm in the north to 1000 mm on Mount Hermon. Based on the climatic index (CI = Precipitation / evapotranspiration) Lebanon can divided into the following Zones (MOA 2003).

Arid 0.10 < CI < 0.25 Dry Subhumid 0.25 < CI < 0.65 Semi Arid 0.65 < CI < 1 Subhumid and humid CI > 1

The zonation of Lebanon according to the Climatic index is indicated in figure 2 below. The main forest and OWL species of Lebanon are Quercus calliprinos, Q. infectoria, Q. cerris, Juniperus excelsea, Cedrus libani, Abies cilicica, Pinus pinea, Pinus brutia, and Cypressus sempervirens. The bulk of the forest area consists of oak and pine stands (ASMAR, F.R., 2003).

1 FAOSTAT area of Lebanon is 10400 km2

Figure 1: Location map of Lebanon

16

Figure 2: Climatic indices of Lebanon according to (MOA 2003) with permanent sample sites of TCP/LEB/2903 indicated.

17

Figure 3: Topography of Lebanon based on Topographic Map of Lebanon 1:20.000 (MINISTÉRE DE LA DEFENSE NATIONALE, 1963)

18

2.2 Existing datasets Prior to TCP/LEB/2903 two datasets related to the forest resources of Lebanon existed (MOA/FAO 2003): 2.2.1 The 1962-1965 1:50.000 Forest Type Map of Lebanon. This data set is prepared from 1962 aerial photography at 1:25.000 with substantial ground checking (DIRECTORATE OF GEOGRAPHIC AFFAIRS, LEBANESE ARMY, 1965). The dataset is of very high quality but as the last 4 decades have seen major changes in the land cover of Lebanon, the 1962-1965 dataset is somewhat outdated. The dataset distinguishes between the following forest types:

Table 1: Forest types of the 1962-1965 Forest Type Map of Lebanon

Forest Type Crown closure Species Oak coppice – open 10 – 30 %

Oak coppice – closed > 30%

Quercus calliprinos, Q. infectoria, with or without some Pinus brutia, Juniperus and Maquis spp. in varying proportions.

Oak standards – open 10 – 30 % Oak standards – closed > 30%

Quercus calliprinos, Q. infectoria, Q. brantii, Q. cerris

Pinus brutia – open 10 – 40 % Pinus brutia – closed > 40%

Pinus brutia

Pinus pinea – predominantly regeneration or pole stands

Pinus pinea – open 10 – 40 % Pinus pinea – closed > 40%

Pinus pinea

Cedar – open 10 – 40 %

Cedar – closed > 40%

Cedrus libani with or without some Quercus spp., Juniperus spp. and Abies silicica in varying proportions

Fir – open 10 – 40% Abies cilicica and Juniperus spp. with or without Cedrus Libani, P. brutia and Quercus spp. in varying proportions.

Juniper – open 10 – 30 % Juniper – closed > 30%

Juniperus excelsea, J. foetidissima with or without Quercus calliprinos and Q. infectoria in varying proportions

Cypress Cypressus sempervirens, usually in mixture with P. brutia.

Areas of mean crown closure <10% i.e. scattered trees and / or dispersed stands too small to map individually are also indicated in the forest type map.

Source: DIRECTORATE OF GEOGRAPHIC AFFAIRS, LEBANESE ARMY, 1965.

19

2.2.2 The 2000-2002 mapping of the Land Cover / Land Use (LCLU) under TCP/LEB/2801. The mapping of the land cover was designed for agricultural census and is based on FAO's Land Cover Classification System (LCCS). The mapping is very accurate and employs minimum polygon-areas of 2000 m2. The merged set of Landsat 5 and IRS-1D described in Chapter 4 were used to produce the map supported by extensive ground validation. The resolution of the merged images is about 5 meters.

As far as the areas of forest and OWL are concerned the LCCS does not translate readily into the global classes employed by FRA because different threshold levels for stand height and crown cover are employed to distinguish forests and OWL from other land. The minimum area employed for the LCLUP is 0.2 ha, while it is 0.5 ha for FRA. Being based on remote sensing with some ground verification, the LCLU map does not contain stand parameters, related to standing volume, uses and management (DALSGAARD, S., 2003A). By combining the information on classification of the LCLU classes used for Forest and OWL in the Land Use Land Cover Project (LULCP) (EL-KHOURY, D. & BAKHOS, W., 2002), with the Standard Description of the FAO Land Cover Classification (FAO 2000), the LCLUP classes used for areas of Forest and OWL as defined by FRA can be summarized as below:

20

Table 2: Summary of criteria for species composition, height and density of LCLU classes containing Forest and OWL

Level 1 Level 2 Level 3 Level 4 Height Crown cover

311 Dense Coniferous Woodland 311a Dense Pines. Pinus spp. mainly P. pinea and P. brutia 311b Dense Cedars. Cedrus libani 311d Dense Fir 311e Dense Cypress. Cupressus spp.

312 Dense Broadleaved Woodland 312a Dense Oaks. Quercus spp 312b Dense - other types of broadleaved trees (poplar, willow)

310 Dense Woodland The crown cover exceeds 60%

313 Dense Mixed Woodland

310/ 112c Urban Sprawl on Dense Woodland

3 – 30 m

> 60-70%

321 Open Coniferous Woodland 321a Open Pines 321b Open Cedars 321c Open Juniper Juniperus spp 321d Open Fir. Abies cilicica 321e Open Cypress

322

Open Broadleaved Woodland 322a Open Oaks 322b Open - other types of broadleaved trees

320 Open Woodland The crown cover is less than 60%

323 Open Mixed Woodland

320/112c Urban Sprawl on Open Woodland

3 – 30 meter

> 10-20% < 60-70%

331 Scrubland 330 Scrubland includes scrubs and other types of degenerated woodland and maquis 332 Scrubland with some dispersed

bigger trees

330/112c Urban Sprawl on Scrubland

300 Woodland Homogeneous zone with a dominant tree or shrub type

340 Burnt woodland

0.3 – 5 meter

> 10-20% < 60-70%

(Source: LICHAA, EL-KHOURY, D. & BAKHOS, W., 2002; FAO 2000).

21

In addition to the classes containing Forest and OWL, the following classes from LCLUP contain woody species and can be defined as Other Land with Trees in accordance with FRA Working Paper 82 (FAO 2004B)

Table 3: Summary of classes from LCLUP that contain Other Land with Trees

Level 1 Level 2 Level 3 Level 4

221 Olives, include both monocultures of dense olive plantations and open olive trees incorporated with arable crops.

223 Fruit Trees include irrigated and dry deciduous fruit trees, main apples, almonds, pears and peaches.

223/221

Permanent crops on field crops.

200 Agricultural area

220 Permanent Crops

224 Citrus Fruit Trees, lemon trees, orange trees.

(Source: LICHAA, EL-KHOURY, D. & BAKHOS, W., 2002; FAO 2000

22

2.3 Justification As a consequence of the 1975 - 1992 civil war, developmental programmes were brought to a stand still for 2 decades. The environmental and natural resource base of Lebanon has been badly degraded and is under serious treat from causing factors such as urban sprawl, overgrazing, quarrying etc. Following the end of hostilities, the resource depletion has been fuelled by the growth in economy and population. Pulling in the opposite direction are changes in employment patterns and migration to urban areas that in some areas have caused abandonment of agricultural areas and grazing lands with subsequent reinvasion of forest. Timely, easily available and accurate information on forest and TOF resources and their utilization is a precondition for sustainable management of these resources based on an environmentally, socially and economically balanced forest policy. The pre-TCP/LEB/2903 level (extent and quality) of information on the forest and TOF resource of Lebanon was perceived inadequate to support national policy decisions and to monitor and counteract resources depletion and environmental degradation. The Government of Lebanon therefore requested the technical assistance of FAO under the Technical Cooperation Programme to assist the Directorate of Rural Development and Natural Resources (DRDNR) to develop capabilities to survey the country’s resources to monitor trends and changes, to describe their nature and localities and to advise on protective measures where required for resources at risk (MOA/FAO 2003). A broad knowledge on the forest and TOF resources is critically needed for redefining the policy and strategy of the Forest Department as well as for developing a comprehensive National Forestry Action Plan (NFAP). The Government considers the formulation of its NFAP as a high priority after the work on the current inventory is concluded.

23

3. Objectives of the Study. According to the Project Document of TCP/LEB/2903, the main objective was to reinforce the capacity of the Directorate of Rural Development and Natural Resources (DRDNR) in collecting, compiling, analyzing and disseminating reliable and up–to-date information on the forest and trees outside the forest (TOF) resources of Lebanon through training of the national staff on forest and tree inventory. To meet this target six main project outputs are described in the Project Document (MOA/FAO 2003): Output 1 The capacity of the Forestry Department of the DRDNR to plan and implement forest

inventories, monitor the resources, manage the related information, and contribute to advance sustainable forest and tree management by enabling an increased use of forestry knowledge in forest policy development and implementation enhanced and strengthened. (For the capacity building aspects of TCP/LEB/2903 please refer to chapter 15)

Output 2 The national team within the Forestry Department and the Regional Development Services adequately trained in forest inventory and assessments techniques and project management through on-the-job training, workshops and study tours. (For the training aspects of TCP/LEB/2903 please refer to section 7.2)

Output 3 As part of the training programme, a forest and tree cover map produced at appropriate scale on the basis of harmonised and standardised vegetation classification system according to national and international requirements. The satellite Landsat TM data available with the UTF/LEB/016 within the Ministry of Agriculture will be used. (For the mapping aspects of TCP/LEB/2903 please refer to section 7.3)

Output 4 Methodology of forest and tree assessment defined on the basis of the approach developed by FRA and taking into account the information needs for national use and international reporting, pilot assessment carried out in selected and representative field sample sites in the country with focus on the multiple functions (environmental, social and economic) of the forest and tree resources, their management, uses and users, health and monitoring of the forest and tree resources set up and a register of permanent observation sites for future surveys established. (For aspects related to development of methodology of TCP/LEB/2903 please refer to section 7.1)

Output 5 A data base based on the pilot assessment on the forestry resources established and the results disseminated to users. (For aspects related to establishing the database for TCP/LEB/2903 please refer to Chapter 9)

Output 6 Priority areas identified and recommendations given to the trained national team from the Forestry Department to build on the project findings in order to develop a forestry action plan, reformulate forestry policy/strategies if needed, identify specific projects for detailed forest inventories, forest and tree resources development and/or conservation, etc. (For the identification of priority areas please refer to chapter 18)

24

4. Resources for the implementation of the NFA. 4.1 Institutions The National Forest and TOF Assessment was undertaken by the Directorate of Rural Development and Natural Resources (DRDNR) and was implemented in the field by the Rural Development Service (RDS), in collaboration with the Forestry Department of the Food and Agriculture Organization of the United Nations (FAO). It was implemented as a project (TCP/LEB/2903) under the Technical Cooperation Programme of FAO. The organization of the project and the allocation of responsibilities is described in Chapter 6 – Project Management. Among the other institutions and organizations that have contributed significantly to the progress of the NFA specifically the following should be mentioned: The Ministry of Defence Cartographic Department provided the project with digital georeferenced colour topographic maps (1:20.000) allowing for the production of accurately georeferenced field maps for the field teams. The Cartographic office also assisted with the scanning and georeferencing of the 1962-1965 map of forest types as well as printing the Derived Forest Map which is one of the main project outputs. The Ministry of Defence, Demining Office provided the maps of the known mined areas – to be used along with the field maps and to identify inaccessible areas (mined areas or areas that for other military reasons are inaccessible). A formalised collaboration between the DRDNR and the Directorate of Studies and Coordination (DSC) within the Ministry of Agriculture concerning the use of and access to data and equipment was encouraged by the Project Document. There was full access for TCP/LEB/2903 to use the merged satellite images, the LCLU map and the A0 plotter of the DSC that were purchased for the Land Cover Land Use (LCLU) Mapping Project (TCP/LEB/2801). The National Council of Scientific Research (NCSR) has made recent satellite images available to the project and been supportive in the original analysis of the LCLU map. The following institutions and organizations contributed to the project through their appointment of professionals and scholars to the Project Steering Committee: Université St. Joseph, National Council of Scientific Research, Lebanese National University, Directorate of Studies and Coordination, Ministry of Environment, Association for Forest Development and Conservation (AFDC - an environmental NGO).

25

4.2 Infrastructure In accordance with the Project Document, the DRDNR provided all physical facilities for the project (office facilities for staff and information system, training space, local transportation including for the fieldwork, means of communications means, etc.). Lebanon is composed of 6 administrative subnational units - the Mohafazas i.e.: Beirut, Mt. Lebanon, North Lebanon, Beqaa, South Lebanon and Nabatiyeh. As no tracts fell in Beirut, the fieldwork was performed in the 5 latter Mohafazas. In each Mohafaza, the staff of the Natural Resources and Rural Development Department (NRRDD) under the Regional Development Services (RDS) under MOA/Regional Services undertook the fieldwork. Each RDS is equipped with several vehicles as is the Central Office of the Ministry of Agriculture – transportation therefore went smoothly. The infrastructure of Lebanon is good. Communications have been possible almost throughout the country via cellular phones except in a few areas of very steep slopes where there was poor coverage. The steep slopes would in some places also cause poor reception for the GPS receivers. The small size of the country combined with a dense network of roads has made it possible for the field crews to return home after each day in the field. In most tracts it has been possible to drive into the tract. In spite of this easy access, the fieldwork in the tracts could still be rather time consuming, and in some plots even impossible, due to the topography and dense vegetation. 4.3 Staff 4.3.1 MOA Staff In accordance with the Project Document, the DRDNR provided all the needed national counterpart staff at secretariat and professional level. The DRDNR through the RDS provided all the national personnel for the field work. The Government appointed a National Project Coordinator, who held the overall responsibility for all aspects of the project activities, with direct reference to the Director General of MOA. Below the national staff at secretariat and professional level are mentioned: Ministry of Agriculture National Project Coordinator: Mr. Ghattas Akl, Director DRDNR. Project Team: - Supervision and Control: Mr. Michel Bassil, forest engineer, DRDNR - Coordination and Assistance: Ms. Zeina Tamim, agricultural engineer, DRDNR - GIS and mapping: Mr. Selim Roukoz, agricultural Engineer, DRDNR - Data Analysis and statistics: Ms. Genane Beydoun, expert; statistics & data analysis Regional Development Services (RDS) / Natural Resources and Rural Development Department (NRRDD) 5 field teams were established, equipped and trained during the first phase of the project. Each field team is headed by the Team leader who is normally the Chief of the NRRDD under MOA/Regional Services in the respective Mohafazas.

26

There was a team leader responsible for all aspects related to data collection in Mt. Lebanon and North Lebanon Mohafaza. Beqaa Mohafaza was split into a northern and a southern part due to the large size and a field team was established for each. The data collection in the two Southern Mohafazas was undertaken by one field team as they are small. On an average this setup implied that each field team had to measure around 45 tracts (BASSIL 2004). A field team would normally consist of 1 team leader and 3 forest guards. An extra number forest guards were trained in the field work – to increase the flexibility, and decrease the vulnerability of the field crews. Since many were trained it was always possible to assemble complete field teams (DALSGAARD, S. 2004A). 4.3.2 External Staff. In accordance with the Project Document, the following consultants / FAO Staff support was attached to the Project. For their respective TORs, please refer to the Project Document:

Table 4: Consultants and external support to TCP/LEB/2903 according to Project Document

Consultant Duration of consultancy

International Consultant Forest inventory and Mapping (IC) 3 months TCDC Forest Inventory Consultant 4 months National Consultant Forest Inventory (NC-FI) 7 months National Consultant Remote Sensing and Mapping (NC-RS&M) 3 months total National Consultant Institutional Development (NC-ID) 4 months FAO backstopping 1,5 months

In addition to the TORs contained in the Project Document, the following additional assistance was recruited during the third and final project phase to assist in the final phases of data entry and the data-processing and statistical analysis.

Table 5: Additional assistance recruited for TCP/LEB/2903

Title Duration of consultancy

Data Entry Service 2 months total Specialist Data Analysis and Statistics 3 months

At the start of the third phase of the project, the recruitment of a specialist in data analysis and statistics was unanimously recommended by the NPC, the IC, the FAO technical backstopping officer and FAO Lebanon to ensure high quality data processing and statistical analysis of the large quantity of detailed field data collected during the field work phase. The specialist has through the data-processing and analysis work ensured that the staff of the DRDNR was adequately trained in data processing and statistical analysis so the staff of the Directorate can perform more tasks related to data processing and analysis in future repetitions of the inventory. Thus the recruitment of the specialist has implied a substantial capacity building component for the DRDNR. For specifics on the organization of the project and the allocation of responsibilities please refer to Chapter 5 – Project Management.

27

4.4. Equipment Wherever possible, equipment was purchased locally to ease any future servicing. The forest measuring equipment however had to be purchased from international suppliers, as it was not available from Lebanese suppliers. Each field team was provided with the following equipment2. An extra set of field equipment was purchased for the supervisory function of the NC-FI and the TCDC expert: 4.4.1 Field equipment Items that were purchased from international suppliers of forestry equipment:

• Fibreglass diameter measuring tape 5 meters • Forest measuring tapes 50 m on V-frame • Measuring tapes 10 m • Prismatic Pocket Compass Model Suuntto • Flagging Tags (rolls) biodegradable • Altimeter Haga • Loggers measuring tape, length 30 m. Spencer • Measuring tape holder • Laser rangefinder. Yardage Pro 1000 • Metal locator Fisher FX-3

Items available for purchasing locally:

• GPS receiver 12 parallel channel receiver band receiver Carmin Etrex VISTA • Digital Camera Nikon 3100 Coolpix • Extra storage media for Digital camera Compact flash TM (256 MB) • Back pack and belt • Galvanized iron poles for marking plot starting point • Binoculars Nikon Standard Action Series 8x40 CF • Slasher and hammers • 4 m poles and flags • Target wooden plate mounted on pole for Laser rangefinder

The field teams were already sufficiently equipped with uniforms and boots, therefore these items were not purchased by the project. The purchase of increment-borers was omitted, as none of the parameters that were included in the survey, after all national recommendations had been added to the FAO base requirements, called for the use of increment borers. 4.4.2 Hardware - Forest Information System: All hardware items were purchased locally:

• 1 workstation with 21” monitor • DVD-RW External • USB memory stick for data transfer • A3/A4 printer • Photocopier A3/A4 with page sorter, stand and printer card. • 2 UPS’s

2 Appendix 6 of DALSGAARD 2003A contains the full technical specifications

28

3 upgraded used PCs were donated to the project from FAO Lebanon in connection with routine upgrading of the hardware at the FAO representation in Lebanon. The PCs were installed at the Rural Development Services and used for decentralised data entry. An A0 plotter was available to the project through the collaboration with DSC and therefore it was not necessary to purchase one for TCP/LEB/2903. 4.4.3 Software - Forest Information System:

• Map Source version 3.02 The following software was already available in the Ministry of Agriculture and therefore not purchased: MS Xp Office Pack Pro, Arc view 3.2. Data analysis was performed by generating the tables of the collected data using different queries in MS Access. After conversion to Excel format the data was analyzed using SPSS3 (BEYDOUN & STEPHAN 2005). 4.4.4 Digital Maps and satellite images: The Ministry of Defence provided the project with: • Digital georeferenced topographic map 1:20000 of Lebanon (1962) for use in the field maps4. • Digital maps of all registered mines and unexploded ordnance for use with the field maps. • Scanned and georeferenced version of the 1962-1965 map of Forest Types of Lebanon.

The Directorate of Studies and Coordination MOA provided the project with: • The digital Land Use Land Cover Map from TCP/2801 • The satellite images available through TCP/LEB/2801 are a merged set of Landsat 5 and IRS-

1D. The resolution of the merged images is about 5 meters as presented below:

Table 6: Satellite images available through TCP/LEB/2801

Sensor Date Band Spatial Resolution Radiometric Resolution Landsat TM 5 October 15, 1998 7 30 m 8 bit

4 30 m 8 bit 2 30 m 8 bit

IRS-1D October 15, 1998 Pan 5 m 6 bit Source: Lichaa, El-Khoury, D. & Bakhos, W., 2002.

3 SPSS: Statistical Package for the Social Sciences is a data management and analysis product for statistical data analysis, including descriptive statistics such as plots, frequencies, charts, and lists, as well as sophisticated inferential and multivariate statistical procedures. 4 An updated set of topographic maps covering the whole country was being prepared by the Cartographic Department of the Ministry of Defence these however were not ready for use during the active period of TCP/LEB/2903

29

As auxiliary data used for the production of the derived forest cover map. The following maps were used: • Map of the forest types of Lebanon 1965 (DIRECTORATE OF GEOGRAPHIC AFFAIRS, LEBANESE

ARMY, 1965). • Map of climatic indices of Lebanon (MOA 2003) • Merged Satellite image of high resolution (Landsat-IRS) • Recent road maps

30

5. Project Management. TCP/LEB/2903 was based in the DRDNR in MOA under the leadership of the Director as National Project Coordinator. The responsibility for the project implementation was allocated to the Department of Forests within DRDNR with the assistance of international and national consultants as well as technical support from FAO-FORM. The organisational diagram for TCP/LEB/2903 is indicated below

National Project Coordinator Mr. Ghattas Akl

Director DRDNR

FAO technical backstopping

Project Steering Committee International Consultant

TCDC expert Expert statistics & data analysis

National Consultant

Institutional Development Ms. Z. Tamim

National Consultant

Forest Inventory Mr. M. Bassil

National Consultant

Remote Sensing and Mapping S. Roukoz / D. Lichaa

Team Leader N. Lebanon

Mr. G.Kassar

Team Leader Mt. Lebanon

Mr. J. Stephan

Team Leader S. Lebanon

& Nabatiyeh Mr. F. Husseini

Mr. Samer.el Khawand

Team Leader North Beqaa

Mr. A. Kozah

Team Leader South Beqaa

Mr. I. Hawi

Field Crew 3 members

Field Crew 3 members

Field Crew 3 members

Field Crew 3 members

Field Crew 3 members

Figure 4: Organisational Diagram for TCP/LEB/2903

5.1 National Project Coordinator As National Project Coordinator to the Project, the Government of Lebanon appointed the Director of DRDNR Mr. Ghattas Akl. With direct reference to the Director General of MOA, the NPC was responsible for the overall supervision and coordination of all project activities, including training of national personnel, forest & TOF mapping activities and organizing the field crews composed of staff from the Regional Development Services.

31

The NPC held the overall responsibility for:

• Coordinated interventions of the national institutions and individuals involved in the project • Facilitation of training of field crew members • Overseeing the fieldwork and securing the timely deployment of the necessary logistical

support of the field crews • Overseeing the design and development of the forest information system and processing

field data and reporting of project findings. 5.2 National Consultants and Team Leaders The main tasks of the National Consultants are mentioned below5. The National Consultant Forest Inventory (NC-FI) was responsible for:

• the training of team leaders and field teams. • the coordination of the fieldwork activities in order to ensure the daily smooth running of

the project. • planning project activities, supervision the field teams and provision of technical guidance

to homogenise the collection of data and the interpretation of variables and definitions. • in collaboration with the TCDC expert, to double check all submitted field forms and

correcting any errors before clearing the field forms for entry into the Forest Information system by the team leaders.

• double checking the structures and contents of the secondary databases filled out by the team leaders before merging and processing the data.

• participation in developing the forest information system and preparing functions for data-processing

• assist in data analysis and reporting of project findings • prepare project progress report and terminal statement.

The main responsibilities of the National Consultant Institutional Development (NC-ID) were:

• assist in reviewing the mandates of the FD and the RDS as to extend to resources monitoring and periodic assessments, under the direct supervision of the NPC/Director DRDNR

• assist in proposing a new organization of the FD within DRDNR and strengthening of the RDS encompassing the new mission regarding forest and tree monitoring and assessment

• assist in defining human and financial resources needed by the FD and the RDS to implement resources monitoring programme and carry out periodic assessments.

• assist DRDNR in defining implementation procedures for the recommendations of the consultancy and to prepare a report for discussion during the workshop on project findings.

5 For complete TORs please refer to Project Document.

32

The main tasks of the National Consultant Remote Sensing and Mapping (NC-RS&M) were to:

• Produce field maps for the field teams. • Produce derived forest cover map, based on the interpretation of satellite imagery,

georeferenced maps of forest types and climatic data and the ground verification provided by the field work and the subsequent ground validation of the draft map.

• Assist in purchasing, installing and using project equipment • In collaboration with the IC and the DRDNR staff to review and adjust the vegetation

classification system, taking into account the national requirements and the FAO terminology and requirements.

• Ensure that the capacity of DRDNR for conducting reinterpretation of images, aided by all relevant existing data was built.

• In collaboration with the IC to prepare a work plan of the mapping activities and to identify timely inputs from the project and the Government.

• Report on mapping results.

The main tasks of the Specialist in Statistics and Data Analysis were to: • Refine and analyze the data in accordance with the agreed strategy for producing results and

statistical calculations and prepare the workshop presentation and store all findings in an easily retrievable format.

• Ensure that the DRDNR counterpart fully understands all the work-processes related to extracting, sorting and analyzing the collected data so that future repetitions will be possible with the DRDNRs own capacity.

• To report any technical problems related to the data and the Forest and TOF Resources Information System to the DRDNR Counterpart.

• To describe all work performed in the form of a terminal report at the end of the recruitment period – to be submitted to and technically cleared by NC-Forest Inventory.

• Assist in any other tasks under TCP/LEB/2903 at the instruction of the NPC, NC-Forest Inventory

The main tasks of the Data Entry Service were to enter the data from the field forms into the primary Access Database at MOA and double-check the data to ensure consistency and quality. The Team Leaders were responsible for all aspects related to the practical field measurements (logistics, quality, safety etc.). The team leaders were responsible for the correct data capture through measurements and interviews and the correct registration of the data in the field forms. The team leaders were also responsible for entering the data in the secondary Access databases at the RDS and the timely submission of reports and databases. 5.3 International Consultants and FAO Support The International Consultant (IC) functioned under the direct supervision of the FRA Programme/RNER and the technical supervision of the FRA Technical Backstopping Officer. Through periodic fieldings the IC provided technical assistance and support to the FD of DRDNR

33

in training of national personnel, planning and implementing the NFA and setting up the information system. The main tasks were:

• Assist in setting up training of national staff through workshops and courses. • In collaboration with national counterparts of the DF at DRDNR to review and adjust

when necessary inventory methodology, vegetation classification and variables. • Assist in supervising the reinterpretation of existing geographic information to construct

the derived forest cover map. • Prepare and update a plan of project activities with inputs from the project and

Government. • Assist in data analysis, reporting of findings and outline follow-up plan of action.

The TCDC Forest Inventory Consultant (IC) functioned under the direct supervision of the FRA Programme/RNER and the technical supervision of the FRA Technical Backstopping Officer. Through periodic fieldings the IC provided technical assistance and support to the FD of DRDNR in training of national personnel, planning and implementing the NFA and setting up the information system. The main tasks were related to:

• Providing continual supervision of field crews on technical issues as to homogenise data collection and best interpretation of variables.

• Reviewing and adjusting when necessary inventory methodology, vegetation classification and variables in collaboration with national counterparts of the DF at DRDNR

• Assisting in preparing and updating a plan of project activities with inputs from the project and Government

• Assisting in organizing collection of field crews outputs, double-checking of field data before clearing for entry in information system by team leaders.

• Assisting in developing the national forest database. • Assisting in data analysis, reporting of findings and outline follow up plan of action.

The FAO/FRA Backstopping Officer undertook periodic missions in support of the project as foreseen in the workplan and provided technical assistance and guidance on aspects of:

• Forest and tree inventory methodology development including sampling design, classification system harmonization and variables.

• Forest and land use type mapping. • Training of national personnel in forest mapping, forest inventory and information

management • Field survey • Data processing, information system development and reporting • Overall technical supervision of project implementation • Technical editing and clearance of project reports including terminal statement.

34

The FAO representation of Lebanon was throughout the project helpful and supportive in all aspects related to the implementation of the various project phases. All Steering Committee Meetings were held at the FAO Representation 5.4 Project Steering Committee In accordance with the Project Document, the Steering Committee provided the overall supervision of implementation of project activities, facilitated inputs to the project at all phases and functioned to ensure a wider dissemination of results. The Steering Committee was composed of representatives from:

• Directorate of Studies and Coordination, MOA • Ministry of Environment, MOE • National Council of Scientific Research, NCSR • Association for Forest Development and Conservation, AFDC • Lebanese National University, LNU • Université St. Joseph, USJ • FAO representation of Lebanon

35

6. Preparations According to the Project Document TCP/LEB/2903 is characterized by 3 well defined phases:

Phase I July 2003 - November 2003: Preparation, Training and Mapping; Phase II November 2003 - July 2004: Field Survey; Phase III July 2004 - September 2005: Data Processing and Reporting.

Phase I includes all preliminary arrangements and preparatory work related to acquisition of equipment, recruitment of staff production of field manuals and field forms, production of working maps and compiling a letter of understanding with DSC promoting inter-Directorate collaboration concerning access to geographic data and use of mapping equipment from an earlier TCP Project, Land Use Land Cover Mapping Project (TCP/LEB/2801). In accordance with the Project Document Phase I also included training sessions for staff, adjustment of methodology and harmonizing classification system and variables according to national requirements. These Phase I project activities are described in chapter 7. The main preparatory activities were as follows: 6.1 Recruitment of national and international consultants The International Consultant was recruited for entry on duty in July 2003 for a total of 3 months split on 5 separate fieldings at strategic times during the lifespan of the project. Prior to the first fielding the IC was briefed by FORM in the inventory methodology in Rome. The National Consultants were identified during the first phase of the project. However due to unforeseen delays concerning the recruitment procedure they were not formally recruited before early 2004. Although not yet formally recruited, all National Consultants worked hard for the project throughout the first phase of the project. In accordance with the Project Document the National Consultants were initially recruited for the following periods.

National Consultant Forest Inventory 7 months National Consultant Remote Sensing 3 months National Consultant Institutional Development 4 months

The TCDC Forest Inventory Consultant, who was supposed to be fielded during the first phase of the project, was fielded during the second project phase in the spring of 2004. The delay was partly due to the climatic conditions of the winter. Due to this delay, the TCDC Forest Inventory Consultant was fielded for a 4 months period (March - July) during the Phase II of the project – the fieldwork phase, i.e. for a shorter total period than the 7 months prescribed by the Project Document. 6.2 Interpretation of satellite images and map construction During the preparatory project phase accurate field-maps were produced of all tracts. The production of maps started after the sampling intensity had been decided upon.

Once the design and intensity of the sampling grid had been decided, an unforeseen challenge related to the production of working maps was that the geographic coordinates provided by FORM for the chosen sampling intensity needed to be converted to Lambert Conformal Conic Levant

36

(LCC) which is the national projection for Lebanon. The working maps were prepared using the national projection while GPS navigation was done using the geographic coordinates (WGS84) as this eliminated the conversion error.

The field maps for the field teams were produced by Mr. Selim Roukoz of DRDNR. These maps show the tract and immediate surroundings with tracts and plots overlaid the georeferenced topographic maps of Lebanon supported by sections of the satellite image and the LCLU map. The working maps also provide the relevant information concerning coordinates of tract centre and SW corner as well as plot-starting points.

The field maps were supplemented by maps of known mines and unexploded ordnance. The data was provided by the Demining Office of the Ministry of Defence and the maps printed by Mr Selim Roukoz. 6.3 Procurement of forest inventory and information system equipment. The list of equipment that was purchased for TCP/LEB/2903 is mentioned in section 4.4. After clearance from FAO FORM during July 2003 it was established that the procurements could be undertaken directly by FAO-Lebanon. The final procurement lists were sent from FAO Lebanon to potential suppliers for collecting offers. For specialized forest equipment 3 external suppliers were contacted in early August 2003 for delivery in Beirut by October 2003, before the start of the training sessions for the field teams.

It was found necessary to modify the original list of procurements of the Project Document for a number of items. Specifically should be mentioned that the following items were added to the procurement list of the Project Document:

• 6 laser rangefinders with 1000 meters range were added to the equipment list, as it would seem a highly useful tool for certain conditions where visibility is good i.e. high altitudes, bare land and agricultural land. Under Lebanese conditions where the range of vision is often high these proved extremely useful.

• 6 metal locators were added to the equipment list allowing for relocating and re-measuring established plots.

• Extra 256 MB storage medias for the digital cameras – to allow for storing a larger number of images than the standard 16 MB media the cameras were delivered with.

• Large hammers for pounding the metal poles into the ground at plot starting points were purchased for each field team.

• Telescopic 4 meter rods were included for each field team. These facilitated navigating in dense maquis vegetation and doubled for the radius measure of the circular subplot (radius = 3,99m, area = 50m2).

The following items were omitted from the procurement list of the Project Document:

• Increment borers were not purchased, as none of the parameters that were included in the survey, after all national recommendations had been added to the FAO base requirements, called for the use of increment borers

• An A0 plotter was available to the project through the Directorate of studies and Coordination and therefore it was not necessary to purchase one for TCP/LEB/2903.

37

• Access database software was already available through the MS Xp Office Pack Pro office package used by the Ministry of Agriculture (MOA). MOA also had Arc view 3.2 thus limiting the need for GIS software purchase to Map Source version 3.02.

• Uniforms and boots were not purchased, as the forest guards were already well equipped.

3 upgraded used PCs were donated to the project from the FAO representation of Lebanon in connection with routine upgrading of the hardware at FAO. The PCs were installed at the NRRDD and used for decentralised data entry by the team leaders. 6.4 Memorandum of Understanding between DRDNR and DSC. A formalised collaboration between the DRDNR and the Directorate of Studies and Coordination (DSC) within the Ministry of Agriculture concerning the use of and access to data and equipment was encouraged by the Project Document. Although a formal Memorandum of Understanding was not signed, there was full access for TCP/LEB/2903 to use the merged satellite images, the LCLU map and the A0 plotter of the DSC that were products of the Land Cover Land Use (LCLU) Mapping Project (TCP/LEB/2801). The signing of a formal MoU would have been a visionary element in the project and would have institutionalized the inter-Directorate collaboration formalizing the synergetic use of the equipment and data collected in an earlier TCP Project, Land Use Land Cover Mapping Project (TCP/LEB/2801). During 2004, DRDNR physically moved to new office premises shared with, among others, the DSC which makes such collaboration even more feasible.

6.5 Field manuals and field forms. During the initial months of the project, the standard FRA templates for the field manual and the field forms were used. Following the recommendations from the workshop on information needs and variables held 06.08.2004 and the additional suggestions for parameters to be measured, the field manual, the field forms as well as the database application were edited simultaneously during December 2003 so that the Lebanese version of the field manual was available to the field teams by the start of the field work phase, by early 2004. Section F4 of the field manual and field forms was revised slightly during April 2004.

38

7. Methodology. 7.1 Sampling design, classification system and variables As mentioned in chapter 3, Project Output number 4 is to define a methodology for forest and tree assessment on the basis of the approach developed by FRA, taking into account the information needs for national use and international reporting, to carry out pilot assessment in selected and representative field sample sites in the country with focus on the multiple functions (environmental, social and economic) of the forest and tree resources, their management, uses and users, health and monitoring of the forest and tree resources and establish a register of permanent observation sites for future surveys. The current section describes the various methodological aspects of TCP/LEB/2903 which took place during phase I of the project. The strategy adopted by TCP/LEB/2903 for deciding on sampling, classification system and additional variables was to launch two workshops for participants from other government bodies, the environmental research community, NGOs, students and the Regional Development Services (RDS) as well as the staff of DRDNR that would be involved in the work. The workshops were held during the first month of the project (July and August 2003). The first workshop would provide the participants with an overview of the standard FAO approach to NFA and raise the questions of which sampling grid, classification system and additional parameters should be adopted for Lebanon. The second workshop aimed at reaching a consensus on the issues among the workshop participants. The workshops are briefly described below: 7.1.1 Workshop on national forest inventory methodology The workshop was held on the 22-23.07.2003. The workshop had about 30 participants from other government bodies, the environmental research community, NGOs, students and the Regional Development Services (RDS). One day of presentations relating to the FAO support to NFA with special emphasis on Lebanese conditions was followed by one day of fieldwork to illustrate the FAO standard methods of data capture and to make it clear to the workshop participants what was registered by the FAO base parameters and variables. This combination would allow the workshop participants to provide qualified contributions to the workshop on information needs and variables which would provide additional suggestions for parameters and how to measure them, as well as suggestions for a vegetation classification system, standardized and harmonized to fulfil national and international requirements. The full report from the workshop of forest inventory methodology, including list of participants is included in the report from the workshop in Appendix 2 of DALSGAARD 2003A. Between the workshops were 2 days of pre-assessment by DRDNR staff, the IC and FORM technical backstopping officer to North Lebanon and to Mt. Lebanon Mohafazas. These pre-assessment trips were instrumental in evaluating the challenges ahead for the field teams in order to determine a realistic sampling intensity. In total around 15 tracts were evaluated.

39

7.1.2 Workshop on information needs and variables The workshop was held on the 06.08.2003 to gather suggestions for:

• a standardized and harmonized vegetation classification system according to national and international requirements.

• additional information needs for national use and how to measure the variables

The workshop had about 30 participants from other government bodies, the environmental research community, NGOs, students and the Regional Development Services (RDS). To allow for constructive feedback on the topics, the workshop participants had during the July workshop been introduced to the theory and the background of the work of FAO in Forest Resources Assessment (FRA) with emphasis on the FAO support to NFA and special focus on Lebanon, the classifications used by the Land Use Land Cover Project and the definitions employed by FAO/FRA. The workshop contained two initial presentations on

1.) the standardized and harmonized vegetation classification system and 2.) the additional information needs.

Both presentations were followed by plenary discussions. A discussion on the basic sampling intensity and the question of whether or not to stratify the country allowing for more intensive sampling intensity in regions of high biodiversity took up a large part of the second working session. The discussion lead to the establishment of a working group that convened on the 08.08.2003, 2 days after the workshop to sum up the points brought forward under the various topics during the workshop, to arrive at a model for the standardized harmonized vegetation classification system and to reach a decision on the issue of stratification 7.1.3 Standardized and harmonized vegetation classification system A Land Classification Chart where the vegetation type is classified according to its structure in terms of crown cover percentage of the tree, shrub and grass layer was proposed by the FORM technical backstopping officer. The initial proposal was adopted with the following amendments to suit the national purposes, i.e.:

• inclusion of riparian forest and other forest areas between 0,5 ha and 0,2 ha (minimal FRA area and minimal mapping area according to LCLU project respectively),

• a breakdown of the pine area into Pinus pinea and other pines, • omission of pastures as a class under agricultural land

. The initial proposal was subsequently refined further by FORM and the IC to arrive at the Land Classification Chart presented on the following page which was subsequently used in the field manual.

40

Table 7: Land Use / Forest Type Classification Lebanon

Tree

s

Shr

ubs

Gra

ss

Land Use / Forest Type

>10 %

5 - 10 %

<5 %

>10 %

<10 %

>10 %

Cod

e

Pinus Pinea X FCPP Pines

Other Pines X FCOP

Cedars X FCCL

Juniper X FCJ

Fir X FCF

Cypressus X FCCS

Coniferous

Mixed coniferous * X FMC

Evergreen X FBE

Deciduous X FBD Broadleaved

Mixed * X FMB

Forest

Mixed * X FMCB

With trees (crown cover 5-10%)

X X SC1 Coniferous Shrublands

without trees (crown cover < 5) X X SC2

With trees (crown cover 5-10) X X SBD1

Deciduous Without trees (crown cover < 5%) X X SBD2

With trees (crown cover 5-10%) X X SBE1

Evergreen Without trees (crown cover < 5%) X X

SBE2

With trees (crown cover 5-10%) X X SMB1

Broadleaved Shrublands

Mixed * Without trees (crown cover < 5%) X X SMB2

With trees (crown cover 5-10%) X X SMCB1

Mixed Shrublands Without trees (crown cover < 5%) X X SMCB2

OWL

Grassland With trees X X X GL1

Riparian X RIP Woodlots (0.2 -0.5 ha) Others X OTH

Grassland X X X GL2

Annual crops CL1 Cultivated Land Perennial crops** CL2

Built-up BUP Artificial Area

Non-built up NBUP

Wetlands X X WETL

Other Land

Barren Land X X BARL

Inland Water WATE * Mixed Forest: Is a forest which contains at least 25 % of one component and 75% at most of the other component (Mixes can be between

coniferous and Broadleaved, between Deciduous and Evergreen broadleaved, and between two species of coniferous). ** Perennial crops: Areas of more than 0.5 ha and with a crown cover of more than 10 % containing olives, citrus trees or other fruit trees e.g. apples, almonds, pears and peaches are classified in the data analysis as the FRA 2005 subcategory Other Land With Trees.

41

7.1.4 Sampling design During the 08.08.2004 meeting of the working group it was agreed to establish a tract every 4 min x 4 min latitude and longitude i.e. a tract every 7.4 x 6.15 km throughout Lebanon, which is a sampling intensity of app. 0.04%. The working group arrived at the decision to adapt a sampling intensity of 4 x 4 minutes after analysing several different sampling intensities (2 x 2, 3 x 3, 4 x 4, 6 x 6 minutes and various stratified combinations of them). The analysis was done by producing a dot grid on a transparent A4 overhead sheet and placing the various dot grids over the map of vegetation levels of Lebanon (ABI SALEH, B., SAFI, S., 1999). Supported by the detailed knowledge of the vegetation of Lebanon possessed by the staff of MOA the various alternatives were evaluated with regards to how well they would capture the various types of Forest and TOF resources of Lebanon. The choice of 4 x 4 minutes seemed to be a good compromise between what was achievable within the time and resources available to the project and broad representation of the various vegetation types. It was also agreed to return to the question of whether to employ stratification in the future to allow for more intensive sampling in areas of high biodiversity after the results of the NFA had been processed and analysed. For an overview of the sample grid please refer to the map of Climatic Indices for Lebanon overlaid with the sample grid contained in section 2.1 and the below illustration of the tract, plots and subplots.

42

Figure 5: Sampling design of TCP/LEB/2903

43

7.1.5 Variables Following the recommendations of the workshops the field manual, field forms and the database application were revised to versions applicable to Lebanese conditions allowing for:

• inclusion of the decided Land Classification system and the decided sampling intensity in the field manual.

• measuring all trees above 10 cm Dbh in all land use sections throughout the plots (i.e. omitting the rectangular subplot). This simplified the fieldwork methodology and captured more trees.

• measuring canopy coverage in all Land Use Sections (F5 section A - 92) • filling out field form F5 section B for both forest and OWL to capture data on management

and structure also in OWL. • setting all commercial height to 0 meters by default as this parameter does not apply to the

conditions in Lebanon. There is no forest industry. The observed silvicultural practices e.g. pruning of Pinus pinea or has been performed for increasing cone production not for reasons related to timber production. In the case of cutting for firewood the practice is one where major branches but not the entire tree is cut.

• inclusion of occurrence and exploitation of some important NWFP in the measurements at circular subplot level (F4 Section B – 77c, 77d, 77e). These are Ferula sp., Gundelia, Origanum, Rheum, Salvia sp., Thymus sp.

• inclusion of section for measuring coppice in (F4 section B – 78da, 78db, 78dc), allowing for registration of average stem height, stem diameter and number of stems per unit.

The measurements of biophysical variables and the interviews were conducted in accordance with the guidelines of the Field Manual. 7.2 Training As mentioned in chapter 3, Project Output number 2 states that the national team within the Forestry Department and the Regional Development Services should be adequately trained in forest inventory and assessments techniques and project management through on-the-job training, workshops and study tours. The current section describes the various training aspects of TCP/LEB/2903. 7.2.1 Workshops The initial workshops held during the initial month of the project, in July and August 2003, provided the participants with an insight in how to establish a national forest assessment in accordance with FAO FRA guidelines, the design of the sample units and what data is collected through the base parameters of the FRA approach. The workshops also included a field trip with some practical measuring exercises.

44

7.2.2 Study Tour As part of the training and capacity building of TCP/LEB/2903 three DRDNR officers visited NFA Guatemala during September 2003 to be trained in fieldwork methodology and data entry and processing6. Table 8: Program for the training of 3 DRDNR engineers in Guatemala 5-14.09.2003 05.09.2003 Arrival to Guatemala 08.09.2003 Introduction to Guatemalan Forestry Sector (Instituto Nacional de Bosques

INAB) Presentation of NFI Guatemala (background and methods) - Visit to Forestry Information System of INAB - NFA Guatemala Organization

09-10.09.2003 Visit to Baja Verapaz to measure a tract (fieldwork) 11-12.09.2003 Workshop on database and data processing 13-14.09.2003 Return to Lebanon

7.2.3 Training of Engineers and Team Leaders The training of the engineers was done during October 2003 with Mr. Michel Bassil NC-FI as the main organizer assisted by Mr. Jean Stephan (Team Leader Mt. Lebanon) and Ghazi Kassar (Team Leader North Lebanon). These three key persons recently returned from the above mentioned study tour to NFA Guatemala and were thus able to share their experiences with the rest of the DRDNR engineers. Ms Zeina Tamim (NC-ID) was essential in organizing the practical aspects. The training sessions were based at MOA Oceanographic Institute in Batroun. The engineers and team leaders were trained in the various measurements, GPS navigation, correct use of measurement equipment and the filling out of the field forms in accordance with the revised Field Manual for Lebanon as indicated below:

Table 9: Administrative distribution of the DRDNR engineers trained in FAO support to NFA and fieldwork methodology

Region

Numbers of DRDNR engineers trained in fieldwork methodology

Beqaa 2

North Lebanon 3

Mt. Lebanon 1

South Lebanon 1

Central FD 3

Total 10 Source (BASSIL 2003)

6 The three officers from DRDNR selected by the NPC for the study tour were Mr. Michel Bassil (NC-FI, DRDNR), Mr. Jean Stephan (Team Leader, Mt. Lebanon RDS) and Mr. Ghazi Kassar (Team Leader, North Lebanon RDS). The host in Guatemala was Ms. Carla Ramirez of NFA Guatemala.

45

The TCDC expert and the NC-FI also facilitated training sessions in data entry for the individual team leaders during May / June 2004 when the final version of the database application was ready and the PCs for decentralised data entry set up at the offices of the team leaders. 7.2.4 Training of Field Teams The training of the forest guards was done during the period October 2003 with Mr. Michel Bassil NC-FI as the main organizer assisted by the other participants of the Guatemala field trip Mr. Jean Stephan (Team Leader Mt. Lebanon) and Ghazi Kassar (Team Leader North Lebanon). The training sessions of the field teams were performed on continuation of the training sessions for the engineers at MOA Oceanographic Institute in Batroun and focussed on training the field crews in the various measurements, GPS navigation, and correct use of measurement equipment in accordance with the Field Manual.

Table 10: Administrative distribution of the DRDNR forest guards trained in fieldwork methodology

Region Number of forest guards trained in fieldwork methodology

Beqaa 17

North 13

Mt. Lebanon 13

South 14

Total 57 Source (BASSIL 2003) According to the Project Document, the formal training of the engineers and forest guards should have been the responsibility of the TCDC-expert. As mentioned in section 6.1, due to an unforeseen delay in fielding the TCDC-expert, the training sessions were not held with the TCDC-expert as the main organizer. When the TCDC-expert (Mr Consolacion of the NFA-Philippines) was fielded in March 2004 he was instrumental, in cooperation with the NC-FI, to ensure constant hands-on training and supervision of the field teams in order to ensure homogenous and accurate field work and measurements by the field teams and in double checking the submitted field reports. The TCDC-expert had great practical experience in undertaking NFA according to the FAO guidelines from the Philippines which was transferred to the field teams through continual on the job training. 7.3 Production of a derived map of forest and TOF resources of Lebanon According to output 3 of the Project Document, a forest and tree cover map produced at appropriate scale on the basis of harmonised and standardised vegetation classification system according to national and international requirements must be produced as part of the training programme. The satellite Landsat TM data available with the UTF/LEB/016 within the Ministry of Agriculture was used. The current section describes the aspects of producing the above mentioned map within the framework of TCP/LEB/2903.

46

7.3.1 Methodology for mapping As mentioned in Chapter 2 the thresholds for crown cover and height of the woody vegetation differ between the classification system employed by the LCLU project and the classification system employed by FRA. An automatic transformation of the polygons from LCLU classes to the classification system employed by TCP/LEB/2903 was therefore not possible. The mapping units of the LULC map are very accurate and employ a minimum size of 2000 m2. The LCLU map was therefore a highly useful base material for the delineation of the forest and TOF resources of Lebanon. There are a large number of LCLU polygons where the translation from LCLU classification to the standardized vegetation classification system of TCP/LEB/2903 is evident and where they can be adopted directly from the LCLU map into one distinct TCP/LEB/2903 class. This is e.g. the case for the Non-Forest and Non-OWL areas such as agricultural areas, urban areas, inland water etc. For the remainder of the LCLU polygons, that all lie in the global classes Forest and OWL, the reinterpretation was done by overlaying the spatial database of the 2002 LULC map (referenced 1998) with earlier georeferenced forest, vegetation maps and satellite images assisted by ground data from the field surveys as well as subsequent ground validation of the draft derived forest map. The existing satellite images were reinterpreted and classified according to the standardized vegetation classification system of TCP/LEB/2903. The reinterpretation of the satellite images was undertaken by the joint National Consultancy in Remote Sensing and Mapping by Mr. Dany Lichaa (who undertook the LCLU mapping under TCP/LEB/2801) supported by Mr. Salim Roukoz of DRDNR. The map production therefore implied a substantial capacity building component for the DRDNR. The following sources of digital georeferenced information were used in the production of the draft derived forest cover map7. • Forest Maps, 1/50.000 (1965)

Source: Ministry of Agriculture & Cartographic Department/Lebanese Army

• Topographic maps, 1/20.000 (1965) Source: Cartographic Department/Lebanese Army

• Satellite images- 5 m resolution (TM merged with IRS) (1998) Source : Agricultural Census Project/ MOA

• Land use/land cover map, 1/20.000 (2002) Source : Agricultural Census Project/ MOA

• Vegetation map of Lebanon, 1/200.000 Source : Lebanese University (Faculty of Sciences)

7 The Cartographic Department of the Lebanese Army had been very helpful in scanning and georeferencing the existing maps. Dr. Samir Safi and Dr. Bernadette Abisaleh of the Lebanese University had kindly provided the project with the Vegetation map of Lebanon for use in the re-interpretation.

47

During February and March 2005, the draft derived forest map was subject to intensive field verification by the NC-FI and the NC-RS&M. The total number of Forest and OWL polygons is app. 12600 (Roukoz 2005). 7.3.2 Output of Mapping The derived forest map was produced by reinterpreting all polygons the Land Cover Land Use Map (referenced 1998) according to the vegetation classification system adopted by the project with support of best available existing georeferenced information concerning vegetation. It is therefore not a new forest map, but a derived map based on existing information supported by new information collected through the fieldwork of TCP/LEB/2903 and the field validation. The derived forest and TOF map of Lebanon is shown in figure 6. The layout of the Derived Map of Forest and TOF resources of Lebanon was prepared for 1/20.000 and 1/200.000 scales in collaboration with the Directorate of Geographical Affairs of the Lebanese Army. Printing layout for A0 (80 x 110 cm) colour format was prepared and maps on 1/200.000 scale were printed. The layout contains information on different forest classes and areas, village locations, road network, rivers, and contour lines (Roukoz 2005). The corresponding area data is shown in table 11 in the results chapter. 3000 maps 1:200.000 have been printed. CD ROMs with the map – as MapInfo file and as an image file have been produced.

48

Figure 6: Derived Forest and TOF map of Lebanon

49

8. Fieldwork and coordination As mentioned in Chapter 6, TCP/LEB/2903 is characterized by 3 well defined phases: Phase I July 2003-November 2003: Preparation, Training and Mapping; Phase II November 2003-July 2004: Field Survey; Phase III July 2004-September 2005: Data Processing and Reporting. The current chapter describes the activities of phase II of the project. In accordance with the Project Document, phase II included the establishment of the permanent sample plots, measurements, observations, interviews and data entry in accordance with the revised Field Manual. 8.1 Distribution of tracts to Mohafazas. The permanent sampling grid is based on a tracts located every 4 x 4 minutes longitude and latitude. This implies that 222 potential tracts fall inside the national land-territory of Lebanon after applying the coordinates to the geo-referenced maps available to DRDNR. The original coordinates provided by FAO/FORM was based on rough delineations and therefore included 226 tracts as the first estimate8. After having received the maps of known mines and unexploded ordnance from the Demining Office of the Ministry of Defence 8 tracts could be excluded as they were mined or otherwise inaccessible. Furthermore in a number of tracts, only some of the plots could be measured due to mines and unexploded ordnance. The distribution of tracts to the individual Field Teams is indicated in the table below and in the map of Lebanon in figure 7 on the following page.

Table 11: Distribution of tracts to the field teams. Region Total Number

of Tracts Inaccessible tracts Number of tracts

for field survey

Mt Lebanon 43 0 43

South Lebanon 44 7 37 South Beqaa 46 0 46

North Beqaa 46 0 46

North Lebanon 43 1 42

Total 222 8 214

Source (BASSIL 2004) 8 Subsequent analysis of the sampling grid has revealed that there is an additional tract in the South East of Lebanon which should be included in future inventories.

50

Figure 7: Distribution of Tracts to Mohafazas

51

8.2 Fieldwork timing and supervision. Fieldwork started late 2003 after the team leaders and crews had received training in the methodology. The field teams finished the work by mid September 2004. The reasons why the 2nd project phase took 9 months rather than 7, as prescribed by the Project Document, are the following:

• the topography and the density of vegetation caused some plots to be time consuming; • the progress of the fieldwork was slower during winter; • the TCDC expert was not fielded before March 2004, (inter alia for climatic reasons); • there was a delay in obtaining the final version of the database from FORM; • the PCs for decentralized data-entry were installed at the RDS during the summer of 2004; • the Team Leaders had other duties for MOA during the fieldwork phase and; • the start of the fieldwork phase fell during the winter season.

Throughout the fieldwork phase, the NC-FI was active in supervising the work and providing guidance on the technical issues related to the measurements, navigation and registration of the data, to ensure homogeneity and high quality in the fieldwork. This function was supported by the TCDC-expert during his fielding March – July 2004. TCP/LEB/2903 did not operate with a control team – instead the continual supervision by the NC-FI and the TCDC-expert functioned to ensure quality and homogeneity in the data capture. The Team Leaders were responsible for all aspects related to the practical field measurements (logistics, quality, safety etc.) as well as for the correct registration of results and the subsequent data entry. After completing the field forms in a tract, the team leaders were responsible for submitting the filled out field forms to the NC-FI/TCDC expert for review. The NC-ID and the TCDC-expert both had to clear the field forms before they could be entered into the database by the team leader in the RDS. In case the field forms were not satisfactorily filled out and the NC-FI / TCDC-expert did not clear the forms for data entry, the field forms would be returned to the team leader with instructions on which data did not fulfil the requirements and instructions on how to amend it. When the field forms for a tract were cleared for entering into the database, they would be returned to the Team Leader for entry into the database. In this manner data entry was decentralized and physically undertaken in the RDS by the responsible Team Leader. Due to the late establishment of functional facilities for decentralized data entry, the data entry did not start up before mid 2004. The entry of data was therefore the bottleneck for the 2nd project phase. Additional assistance was therefore recruited to assist in the final phases of data entry. 8.3 Field work experiences Lebanon is a small country with a good infrastructure. It was therefore often possible to drive close to the tract and often into it. It was also possible for the field teams to return home after every day in the field. The current section summarizes the observations from the team leaders presented at the technical workshop in Beirut on the 11th of January 2005.

52

8.3.1 Summary of Team Leaders experiences concerning fieldwork9

• Prepare a monthly work plan of tracts to be done • The monthly plan should include tracts of different administrative units and different levels of

difficulty – to give variation in the work and a visible progress and to avoid long periods of long travel to remote sites.

• Start from sea level and work upwards (since work started in November) • Keep the weekly plan flexible according to weather conditions • Notify director of Regional Service about weekly work plan • Obtain owners permission when needed before starting the work • Fruit trees are normally pruned below 1.3 meters and are therefore alternative methods for the

measurement of these should be investigated. • For better efficiency and less time and fuel consumption, long working days are required (start

early, and leave no later than sunset). • Photo archiving and field form review by team leader should be done later in the evening on

daily basis if possible when the memories are still fresh.

8.3.2 Summary of Team Leaders evaluation of Field forms Field form 1 • Locality name rarely used • Informant addresses are difficult to get • Subjectivity of information concerning altitude of tract center and population • Description of roads through written comment and photos is enough; GPS is not recommended

Field forms 2 & 3 • F 2: References for starting point are sometimes hard to find. Photos are essential • F 3: Plant species identification in some cases of deciduous and ornamental trees is difficult • F 3: Slope correction for each tree is done later, while total length of plot is corrected directly • F 3: Causing agent for tree health problems is hard to define • F 3b: Was non applicable to Lebanese conditions

Field form 4 • Soil drainage classes should be reconfigured differently • Ferula sp., Rhubarbia sp. and Gundelia tournifolii are found outside forests or OWL, and thus

not measured: they could be omitted • Salvia sp., Origanum sp. and Thymus sp. are counted in winter and spring, thus it cannot be

determined if exploited • Coppice estimation is subjective

Field form 5 • Determination of LUS and tree canopy coverage is subjective in mixed vegetation, in

fragmented areas and when tree cover is low, or tree cut • Protection status is mostly production which does not reflect the reality • Shrub coverage can be under trees or between them: percentages cannot be added

9 From STEPHAN 2005

53

• Disturbances are related to environmental problems cited in the form, fire and grazing • Estimation of the surface of fire and frequency is subjective • Timber exploitation and silviculture are absent in most of the cases due to the prohibition of

tree cutting for broadleaved trees for the last decade and the prohibition of conifer felling

Field form 6 • Most products and services can be seen onsite without interviewing, but ranking is only based

on economical importance • User rights and conflicts mainly concern fruit and wood exploitation, grazing and hunting • Demand and supply trends, end use and extraction trends are very subjective • Frequency is mainly seasonal, except for charcoal production • Legislation and compliance only concerns protection, wood exploitation and hunting • Species were not defined for non-wood products and for services

Photos evaluation • Photos are numbered according to tract and plot numbers (e.g. 151-2.1) • Best description of reference points and plot access • Best description of species, land use and cover inside plots, including inaccessible areas • Should be worked on in a photo editor for better accuracy and to add information

8.3.3. Team Leaders evaluation of equipment and fieldwork. Measuring tape-holder and binoculars can be omitted from the equipment list. Some of the measuring tapes broke during fieldwork – stronger versions of these are recommended. A supply of extra equipment should be available to the Field Teams in the case something is damaged or lost. The GPS receiver was inaccurate under dense tree cover and in deep valleys. For determining the plot axis in urban areas and inaccessible areas it was very useful. The laser rangefinder and 3m pole with wood board was very practical for reference definition (over 20m), outside forests, and in inaccessible areas. Finding a metal pole at the starting point could be very time consuming. Great care should be taken to note its location accurately. Some injuries occurred during fieldwork as well as damage to shoes and clothes. A mechanism for replacement of damaged working clothes should be considered. To ensure the best future repetition of the NFA the Team Leaders recommended to; • Review field forms in accordance with the comments in section 8.3.2 for better data objectivity • Clear all mined areas for better accessibility to plots. • Make sure that the tracts are done by the same crew for better efficiency and accurate

monitoring • Make ample resources available to future repetitions of the work in the form of human resources

and logistics (car and fuel + maintenance) and provide incentives for undertaking extra work, within the ministry’s budget.

54

8.4 Coordination of activities The overall responsibility for coordination activities was the responsibility of the National Project Coordinator. The daily coordination of the fieldwork and the data entry was done by the NC-FI. The NC-ID held a secretariat function and handled issues related to purchases, formal contacts to other institutions among these the project Steering Committee and FAO Lebanon. In accordance with the recommendations of the NC-FI and the TCDC-expert, the team leaders were requested at regular intervals to submit progress reports containing a status for the field work and the data entry in the respective Mohafazas as well as a workplan for the remainder of the field work phase. This would facilitate keeping track of the fieldwork progress in order to allow for timely identification of areas in need of consolidation. Formal progress reports were submitted for Mt. Lebanon Mohafaza, the rest were provided verbally and included in the overall progress reports for the field work compiled by the NC-FI and NC ID. The Steering Committee was briefed on the progress of the work every time it convened and provided its overall recommendations on project related issues. The fieldings of the IC, the TCDC-expert and the FORM backstopping Officer were located at strategic periods during the lifespan of the project. The actual timing of the consultancies differed slightly from the workplan of the initial Project Document as the workplan was adjusted a couple of times as the project progressed.

55

9. Database design, data-entry, processing and analysis. According to Output 5 of the Project Document a data base based on the pilot assessment on the forestry resources must be established and the results disseminated to users. The current Chapter Describes all aspects related to database design, data entry, data processing and data analysis. As mentioned earlier TCP/LEB/2903 is characterized by 3 well defined phases. Designing the database took place during the first and the second project phase while data processing and data analysis occurred during the second project phase: 9.1 Database Design and data structure The basic design and structure of the database is the standard database developed by FAO FORM. It is an MS Access application. A version adjusted to fit the Lebanese NFA and its standardized harmonized vegetation classification system and additional national variables was finalized at FAO FORM during the 2nd project phase in early 2004. 9.2 Data entry As mentioned in section 8.2 the team leaders were responsible for submitting the filled out field forms to the NC-FI/TCDC expert for review. After both the NC-FI and the TCDC expert had cleared the submitted field forms for entry, the team leaders were responsible for the data entry on PCs that were installed at the RDS. The individual team leaders had received hands-on training in the respective RDS’s by the NC-FI and the TCDC-expert to ensure that the data entry was performed correctly. Before a set of field forms was cleared for data entry the data needed to be checked and cleared by the NC-FI or the TCDC–expert. Field forms that were found to be incomplete was returned to the responsible team leader with instructions on what was missing or incorrectly registered and how to complete the form. Due to the late establishment of functional facilities (PCs and database application) for decentralized data entry, the data entry did not start up before mid 2004. The entry of data was therefore the bottleneck for the 2nd project phase. Additional assistance was therefore recruited to assist in the final phases of data entry in order to have all field data entered by late September 2004. The data entry assistants were well educated forest guards who had participated in the field work and therefore understood the data. After the team leaders had completed their data entry and submitted the files, the NC-FI double checked the database structures and data consistency before merging the secondary databases from the RDS’s into the main database at MOA. The FAO backstopping officer had through his 2 week mission to Lebanon in September 2004 trained the NC-FI in performing these tasks After merging all the field data into the primary database (the Forest Information System mentioned in the Project Document) the data was ready for processing and analysis.

56

9.3 Data processing The tasks related to refining, processing and analyzing the data are highly specialized. In addition to the TORs contained in the Project document and the data entry assistants mentioned in section 9.2, a specialist in data analysis and statistics was therefore recruited during the third and final project phase to assist in data-processing and statistical analysis for a total of 3 months. The recruitment of the specialist in data analysis and statistics was unanimously recommended by the NPC, the IC, the FAO technical backstopping officer and FAO Lebanon, at the start of the third project phase, to ensure high quality data processing and statistical analysis on the large quantity of detailed field data collected during the field work phase. The specialist has through the data-processing and analysis work ensured that the staff of the DRDNR was trained in data processing and statistical analysis so that the staff of the Directorate can perform tasks related to data processing and analysis in future repetitions of the inventory. Thus the recruitment of the specialist has implied a substantial capacity building component for the DRDNR. 9.4. Data analysis The first step of the analysis procedure was to generate the tables of the collected data using different queries in MS ACCESS. These tables are obtained by classification procedures, based on the definition of each variable. The tables are then converted to Excel format which can be used by the SPSS software10. Once the tables are obtained in the SPSS software, the descriptive statistics and the estimation based on the Ratio analysis method are applied on all measured variables which depend on the�size of the�area over which it’s measured. All functions used for the study are illustrated in appendix 1. Many of, the measured variables depend on the size of the area over which it is measured. For these variables Ratio estimates are used.

10 SPSS: Statistical Package for the Social Sciences (a data management and analysis product for statistical data analysis, including descriptive statistics such as plots, frequencies, charts, and lists, as well as sophisticated inferential and multivariate statistical procedures).

57

10 Results 10.1 Area TCP/LEB/2903 produced two area data sets - one from the systematic field sampling of the NFA and one from the derived forest cover map. In subsections A and B in the below section 10.1.1 the datasets are presented and they are commented in section C. 10.1.1 Area by forest types and Land Use Classes A – Area data extrapolated from NFA Systematic sampling.

Table 12: Area by Global classes (level 1) from NFA Land Use Area Size x % (ha)

Forest 211 13.3 139,376 OWL 211 10.4 108,378 Other Land 211 76.3 797,152 Inland Water 211 ---- 294 Total area 100 1,045,200

(Source: BEYDOUN & STEPHAN 2005)

Table 13: Forest area by forest type (level 2) from NFA Type Size x % (ha)

Coniferous 68 32.2 44,879 Broadleaved 68 56.6 78,887 Mixed forest11 68 11.2 15,610 Total area 100 139,376

(Source: BEYDOUN & STEPHAN 2005)

11 Mixed forest is defined as forest containing at least 25% each of coniferous trees and broadleaved trees

58

Table 14: Area estimates for coniferous and broadleaved forest types (level 3 & 4) from NFA

Forest type size x % (ha)

Coniferous Other pines 29 40.0 17,952 Juniper 29 23.4 10,502 Pinus pinea 29 17.7 7,943 Mixed Coniferous 29 11.6 5,206 Cedars 29 4.5 2,019 Cypressus 29 2.8 1,257 Fir 29 0 0 Total coniferous 100 44,879 Decidous Evergreen 48 41.8 32,975 Mixed

Broadleaved 48 38.8 30,608

Deciduous 48 19.4 15,304 Total area 100 78,887 Mixed forest

68 11.2 15,610

Total area 100 15,610

(Source: BEYDOUN & STEPHAN 2005)

Table 15: OWL area by OWL type (level 2) from NFA Type Size x % (ha)

Broadleaved shrub 69 52.1 56,465 Grassland with trees 69 33.4 36,198 Mixed shrublands 69 13.3 14,414 Coniferous shrubs 69 1.2 1,301 Total 100 108,378

(Source: BEYDOUN & STEPHAN 2005)

Table 16: OL area by OL type (level 2) from NFA OL type Size x % (ha)

Grassland 205 57.6 459,160 Cultivated land area 205 28.1 224,000 Artificial area 205 10.4 82,904 Barren land 205 3 23,914 Woodlots area 205 0.9 7,174 Wetland 205 0 0 Total 100 797,152

(BEYDOUN & STEPHAN 2005)

59

B – Area data from Derived forest map

Table 17: Area data from the derived forest and TOF map of Lebanon

FRA Global Class

Area (ha)

% of total area

Forest types Area (Ha)

% of total area Code

Area (ha) % of total

area

FCCL Coniferous Forest Cedrus libani 2125 0.208

FCCS Coniferous Forest Cupressus sempervirens

319 0.031

FCF Coniferous Forest Abies cilicica 1613 0.158

Coniferous 42649 4.16 FCJ Coniferous Forest Juniperus excelsa 11318 1.105

FCOP Coniferous Forest other pines 14533 1.419

Forest 134372 13.1 FCPP Coniferous Forest Pinus pinea 12740 1.244

61325 5.99 FBD Broadleaved Deciduous Forest 16958 1.656

Broadleaved FBE Broadleaved Evergreen Forest 44367 4.333

FMB Mixed Broadleaved Forest 10005 0.977

Mixed 30398 2.97 FMC Mixed Coniferous Forest 3296 0.322

FMCB Mixed Coniferous & Broadleaved Forest 17097 1.670

120574 11.8 OWL with trees S1 Other Wooded Lands with trees 75505 7.374 Other wooded

land OWL without trees S2 Other Wooded Lands without trees 45069 4.401

766289 74.9 Other land with trees 58 0.006 RIP Riparian Forest 58 0.006 Other

land

Other Land without trees 766231 74.79 OL Other Land without trees 766231 74.79

Inland Water 1745 0.2 17.45 0.17 IW Inland Water 1745 0.17

Source (Roukoz 2005)

60

C – Comparison of datasets from NFA and DFM. The two datasets from TCP/LEB/2903 are compared in figures 8 -10 below

Figure 8: National area (%) by FRA Global Classes from DFM and NFA

0

10

20

30

40

50

60

70

80

90

100

DFM NFA

Per

cent

Inland Water

Other Land

Other Wooded Land

Forest

Figure 9: Forest area by forest type from DFM and NFA

0

20000

40000

60000

80000

100000

120000

140000

160000

DFM NFA

hect

ares

Mixed forest

Broadleaved decidous

Broadleaved mixed

Broadleaved evergreen

Fir

Cypressus

Cedars

Mixed coniferous

Pinus pinea

Juniper

Other pines

61

Figure 10: OWL area by OWL type from DFM and NFA

0

20000

40000

60000

80000

100000

120000

140000

DFM NFA

hect

ares

OWL without trees

OWL with trees

Coniferous shrubs

Mixed shrublands

Grassland with trees

Broadleaved shrub

The datasets show very similar percentages at the global class level / Level 1 (figure 8) The NFA estimates the forest area of Lebanon at 13.3 % of the total area of the country, while the DFM determines the forest area at 13.1 %. The NFA estimates the OWL area of Lebanon at 10.4 % of the total area of the country, while the DFM determines the OWL area at 11.8 %. Concerning the individual forest types (figure 9) the main differences are apparent different interpretations of the categories Broadleaved mixed and Broadleaved evergreen and the fact that the NFA did not capture any fir areas in the sampling grid. Concerning the areas of OWL (figure 10) – the DFM and the NFA employed different classifications. The total OWL areas are similar. A comparison at the level of the individual OWL type is however not possible due to the different classifications employed. During the reinterpretation of the LCLU map it became apparent that it was not possible to distinguish between the various OWL classes employed by the NFA. A different subdivision of OWL has therefore been employed for the mapping of OWL (OWL with trees and OWL without trees). As indicated in table 18 the findings from the DFM and the NFA at level 2 for forest area also compare well. Both indicate that slightly more than half the forest area of Lebanon is broadleaved forest, a third of the forest area is coniferous forest while the remaining 11-13 % is mixed forest.

Table 18: Percentage of forest area by forest type according to DFM and NFA.

Forest Type Level 2 DFM (%) NFA (%) Coniferous Forest 34.2 32.2 Broadleaved Forest 53.1 56.6 Mixed Forest 12.7 11.2 Total 100 100

62

Comments on coniferous forest (BEYDOUN & STEPHAN 2005): Conifer forests are found mainly in the subtropical mountain zone in Mediterranean area. Through the times they have been exploited for wood consumption, and their area is assumed to have decreased dramatically. According to the NFA, 45,000 ha. of coniferous forest exist in Lebanon, primarily pine plantations for production of nuts, protected coniferous stands and reforested areas. The class “Other pines” contains mainly Pinus brutia, and to some extent also Pinus halepensis occupies the largest area. These species thrives in all type of soils up to altitudes of 1500 m. They are pioneer species. They grow and grow fast and spread easily after fires, which explains why they occupy the biggest area, especially in subtropical dry forest ecological zone. Juniper (mostly Juniperus excelsa) comes in the second place because it occupies large areas of the subtropical mountain, above 1800 m where other species do not compete. Pinus pinea is widely planted up to 1500 m mainly on sandy siliceous soils. Mixed forests contain fragmented and scattered stands of mixed pines, mixed pine and cypress and mixed conifers plantations. Cedar groves (Cedrus libani) are few and scattered on the western slopes of the western chain between 1300 and 1950m. They include spontaneous stands, and plantations. This species has been heavily exploited for its wood since antiquity, and is now reduced to scattered stands covering only 2000 ha. Cypress (Cypressus sempervirens) is rarely found in pure stands. It is a pioneer species in marginal lands with steep slopes. It is quickly accompained by Pinus brutia and is therefore mostly found in mixed stands. Fir (Abies cilicica) which is mainly found in pure stands or mixed with cedar and juniper in certain areas of North Lebanon was not measured within the tracts. Comments on deciduous forest (BEYDOUN & STEPHAN 2005): Broadleaved species occupy the largest area because they constitute the main original vegetation formation in the Mediterranean area. In the subtropical dry forest and steppe, they form evergreen and mixed broadleaved coppices. In subtropical mountain they form mixed and deciduous forests. Broadleaved species have always been always, but as they regenerate easily (naturally and by coppicing) after cutting, they still occupy large areas although often in a somewhat degraded state. The proportion of each forest type is correlated to the different ecological zones. Quercus calliprinos which is the dominant species in evergreen forests grows between sea level and 1500 m. Quercus infectoria is found as an accompanying species to Q. calliprinos from 200 m, and contribute the most in mixed forests formation. In the subtropical mountain ecological zone, cool and humid weather associated with a flattened relief and deeper soil make the optimal condition for deciduous forest formations of Quercus infectoria, Quercus cerris, Quercus brantii and Ostrya carpinifolia. Comments on Other Wooded Land (BEYDOUN & STEPHAN 2005): In the Mediterranean area, broadleaved shrubs which regenerate easily by coppicing are found in the subtropical dry forest and a part of the subtropical mountain, this category is therefore the most common. It is followed by the grassland with trees, mainly juniper formations in high altitudes. Mixed shrublands occur mainly between the subtropical mountain and subtropical steppe ecological zones. Coniferous shrubs are rare and consist mainly of degraded juniper formations and dwarf cedar plantations that grow beyond their natural geographical zone, and cannot reach a tree height above 5 m in situ.

63

10.1.2 Area by Ecological Zone According to the NFA, the GEZs take up the following percentages of the national area:

Subtropical mountain system 31.3 % Subtropical dry 42.8 % Subtropical steppe 25.8 %

Table 19: NFA Estimates of forest area by Global Ecological Zone. FRA

Global Class

Global Ecological Zone

Size x % (ha)

Forest Subtropical mountain system 67 48.0 66,901 Subtropical dry 67 38.1 53,102 Subtropical steppe 67 13.9 19,373 Total 100 139,376 OWL Subtropical mountain system 69 44.3 48,011 Subtropical dry 69 31.7 34,356 Subtropical steppe 69 24.0 26,011 Total 100 108,378 OL Subtropical mountain system 205 26.6 212,042 Subtropical dry 205 45.2 360,313 Subtropical steppe 205 28.2 224,797 Total 100 797,152

(Source: BEYDOUN & STEPHAN 2005)

It can therefore be concluded that GEZ Subtropical mountain system is relatively more common in areas classified as Forest and OWL. The relative frequency of GEZ Subtropical dry and Subtropical steppe is relatively more common in areas classified as Other Land.

Half of the forests are located in subtropical mountains, because it is the most suitable ecological zone for forest formation in the Mediterranean area (cedar, fir and deciduous broadleaved forests). This zone is also less exposed to demographic pressure. It also includes a large proportion of the OWL, which is mainly composed of broadleaved coppice and juniper formations. The Subtropical dry forest GEZ on the coastal areas up to 1000 m is characterized by evergreen broadleaved coppice (Mediterranean maquis) which shares 50% of the OWL, and Pinus brutia and Pinus pinea stands that constitute about 40% of the forest cover. Subtropical steppe has a limited share of forest and OWL cover of mainly broadleaved tree species (BEYDOUN & STEPHAN 2005). 10.1.3 Proportion of forest area by management system Forest area by Management system was not analyzed due to the virtual absence of forest management in Lebanon.

64

10.1.4 Proportion of forest area for each protection level The classification of the designation/protection status in Lebanon is composed of the classes Production, Protection and Natural Reserve. According to the NFA the area of the land use classes by designation/ protection status is as follows.

Table 20: Forest area by designation / protection status Designation/protection

Status Size x % (ha)

Production 59 97.4 135,752 Natural reserve 59 2.6 3,624 Protection 59 0 0 Total 100 139,376

(Source: BEYDOUN & STEPHAN 2005) Table 21: OWL area by designation / protection status

Designation/protection status

Size x % (ha)

Production 65 96.9 105,018 Natural reserve 65 3.1 3,360 Protection 65 0 0 Total 100 108,378

(Source: BEYDOUN & STEPHAN 2005)

Table 22: OL area by designation / protection status. Designation/protection

Status Size x % (ha)

Production 203 97.5 777,223 Natural reserve 203 0.4 3,189 Protection 203 0.2 1,594 Not Known 203 1.9 15,146 Total 100 797,152

(Source: BEYDOUN & STEPHAN 2005)

Natural reserves cover only a modest part of forests, OWL and OL. The actual natural reserves are found on state, communal and municipal lands, evenly distributed between forests, OWL and OL and cover almost 10200 ha or 1% of the country. There is no clear definition for the existing protection status in Lebanon and the area figures are there connected with some uncertainty (BEYDOUN & STEPHAN 2005).

65

10.1.5 Proportion of forest area by ownership

Table 23: Forest area by ownership category Land tenure size x % (ha)

Private12 59 60.4 84,183 State 59 27.4 38,189 Municipality 59 10.0 13,938 Community 59 1.2 1,672 Not Known 59 1.0 1,394 Total 100 139,376

(Source: BEYDOUN & STEPHAN 2005)

Table 24: Other Wooded Land by ownership category

Land tenure size x % (ha)

Private 65 80.0 86,702 Public 65 13.8 14,956 Not Known 65 6.2 6,720 Total 100 108,378

(Source: BEYDOUN & STEPHAN 2005)

Table 25: Other Land area by ownership category Land tenure size x % (ha)

Private 203 88.2 703,088 Public 203 11.0 87,687 Not Known 203 0.8 6,377 Total 100 797,152

(Source: BEYDOUN & STEPHAN 2005) It is characteristic that private ownership dominates in all 3 classes. 10.1.6 Precision of area estimates As mentioned at the level of the Global Classes (Level 1), the findings of the Derived Forest Map (DFM) and the NFA compare well. There is also a good comparison between the datasets at Level 2. The differences between the datasets become more pronunced further down in the classification hierarchy at level 3 and 4. This is due to the fact that the NFA is based on systematic sampling of app. 0.04 % of the land area of Lebanon and that the results from the survey have been extrapolated to national level which naturally causes deviations between the two datasets especially for land use types that are rare (e.g. fir and Inland Water).

12 The category Private ownership includes the properties of the church.

66

The area estimates from the derived forest map are based on mapping that covers the entire country. The recommendation is therefore to use the area data from the DFM as it is more accurate. Concerning the area data related to protection level and ownership - the figures from the fieldwork should be seen as indicators only. The figures show a large proportion of all land types are in private ownership and a low proportion is under formal protection. More detailed studies are needed to obtain figures of use for management purposes. For a complete overview of the analysis of the area including Standard Errors please refer to BEYDOUN & STEPHAN 2005 10.2 Volume 10.2.1 Gross Volume of Global Classes Based on the terms and definitions of the FRA 2000, Volume is defined as the stem volume of all living trees more than 10 cm Dbh (or above buttresses if these are higher), over bark measured from stump to top of bole. The estimation of the tree stem volume is obtained by using the following formula:

Tree stem volume (m3) = ��

���

�orVolumeFactHeighttotal

DBH***

4)( 2

π

The Volume Factor is a correction of the tree stem’s cylinder volume. This factor depends on the type of tree and is around 0.5 – 0.7 and is species specific. For fruit trees 0.55 has been applied. The volume per hectare is obtained by the sum of the tree stem volumes. The estimation of the total volume (in m3) is obtained by multiplying the volume per hectare with the total area of that land use in the country. Table 26 below presents the volume figures for the 3 global classes.

Table 26: Estimated volume for global classes for tree with Dbh >10cm. Land use Size Volume/Area

( m3/ha) Total Volume

(m3)

Forest 61 35.64 4,967,361 OWL 65 5.08 550,777 Other land 149 3.74 2,984,537

Total 8,502,675

(Source: BEYDOUN & STEPHAN 2005) The volume data is in the current report is not presented by individual forest types. The Forest information management system however allows for extracting data by forest type.

67

Measurement of commercial volume was been omitted from the survey. There is no forest industry in Lebanon and felling in natural forests is prohibited. Furthermore certain silvicultural practices e.g. cutting branches for firewood or pruning in Pinus pinea for increased cone production, renders the wood without industrial value due to large knots (BEYDOUN & STEPHAN 2005). 10.2.2 Precision of volume estimates The volume figures of the NFA should be considered first estimates, as specific volume functions and form figures are absent and as the volume estimates are base on a 0.04 % sample of the national area. The size of the average growing stock per hectare for all types of forest is 35.6 m3 which seems realistic. In light that this information was virtually lacking prior to TCP/LEB/2903, the NFA represents an improvement in knowledge based on some initial assumptions concerning form factors and with the clear limitations that a 0.04% sample has. Care however must be taken to verify these findings through follow up studies and to conduct further studies and analysis that allows for determining forest type specific average growing stock and specifying them further according to GEZ and altitude. A realistic balance between the need for new information and available resources should be struck. The forestry sector in Lebanon is small which may limit the amount of resources that can be channelled into obtaining figures that are more accurate. As far as volume figures in OWL and OL are concerned they are estimated at 5.1 m3 in OWL and 3.7 m3 in Other Land. These may well be underestimated due to the methodology for measuring trees employed by the FRA approach to NFA where the stem of trees above 10 cm is measured at

1.3 meters above ground. Orchard and olive trees make up the bulk of trees in OWL and OL and they are almost all pruned below 1.3 meters meaning that the measurements are performed on the major branches. As orchard trees are legally available for wood supply through the removal of non-productive trees, it could be an interesting study to further investigate more representative and less labour intensive methods of determining the growing stock in orchards and olives. Figure 11 illustrates the problems related to measuring orchards trees.

Figure 11: The problem of measuring orchard trees. Photo: Ibrahim Hawi, Team Leader Bekaa South

68

10.3 Biomass The methodology used for biomass calculation is based on the following formulas:

SB= GS x WD AGB=SB x BEF BGB=AGB x R

Where SB Stem biomass (tonnes) AGB Above-ground biomass (tonnes) BGB Below-ground biomass (tonnes) GS Growing stock (volume) over bark (m3) WD Wood density (Dry weight/green volume expressed in tonnes/m3) BEF Biomass expansion factor (above ground biomass/stem biomass) R Root-shoot ratio (below-ground biomass/above-ground biomass)

10.3.1 Biomass of forest species For this calculation, only the 10 species with the largest growing stock (volume over bark) were analyzed.

The value of the wood density factor (WD) for coniferous and broadleaved species are assumed to equal 0.4 and 0.5 respectively. For some species a more specific WD factor has been applied (e.g. 0.58 for oaks)13.

For the Biomass expansion factor (BEF), the value of 1.3 is used for the Coniferous, and the value of 1.4 is used for the Broadleaved.

The Root-shoot ratio used is 0.27, as Lebanon is considered the subtropical dry forest. Applying these factors to the data collected through the NFA the following figures for Growing Stock and Above/Below Ground Biomass.

Table 27: Growing Stock and biomass of the 10 most important forest species. Type

GS (million m3)

AGB (million tons)

BGB (million tons)

Quercus cerris 0.96 0.78 0.21 Pinus Pinea 1.55 0.62 0.22 Pinus brutia 1.3 0.68 0.18 Cedrus libani 0.31 0.16 0.04 Juniperus excelsa 0.29 0.15 0.04 Juniperus drupacea 0.09 0.05 0.01 Quercus Infectoria 0.186 0.15 0.04 Quercus calliprinos 0.177 0.14 0.04 Ostrya carpinifolia 0.016 0.01 0.003 Platanus orientalis 0.019 0.013 0.004 Others species 0.06 0.04 0.01 Total 4.967 2.793 0.797

(Source: BEYDOUN & STEPHAN 2005) 13 For details please consult (BEYDOUN & STEPHAN 2005)

69

10.3.2 Biomass of TOF species Similarly the biomass and growing stock for the most common TOF species or species groups are presented in table 28 and 29 below:

Table 28: Growing Stock and biomass of the 10 most important TOF species / species groups in Other Wooded Land Type

GS (million m3)

AGB (million tons)

BGB (million tons)

Juniper Excelsa 0.32 0.17 0.04 Quercus Calliprinos 0.1 0.08 0.02 Pinus Brutia 0.05 0.03 0.01 Pinus Pinea 0.02 0.01 0.003 Pyrus syriaca 0.01 0.01 0.002 Populus alba 0.01 0.005 0.001 Quercus infectoria 0.01 0.008 0.002 Prunus Amygdalis communis 0.006 0.004 0.001 Prunus Ursina 0.006 0.001 0.001 Others species 0.012 0.007 0.002 Total 0.544 0.325 0.082 (Source: BEYDOUN & STEPHAN 2005)

Table 29: Growing Stock and biomass of the 14 most important TOF species / species groups in Other Land Type

GS (million m3)

AGB (million tons)

BGB (million tons)

Olive 0.829 0.58 0.157

Eucalyptus spp. 0.434 0.304 0.082

Casuarina Equisetifolia 0.316 0.221 0.06

Pinus Pinea 0.2 0.104 0.028

Citrus 0.17 0.116 0.03

Cupressus sempervirens 0.107 0.056 0.015

Juniper Excelsa 0.1 0.05 0.01

Quercus Calliprinos 0.1 0.08 0.02

Prunus 0.1 0.07 0.02

Quercus Infectoria 0.07 0.06 0.02

Platanus orientalis 0.06 0.044 0.012

Prunus Amygdalis 0.05 0.04 0.01

Apple 0.047 0.033 0.009

Others species 0.401 0.21 0.056

Total 2.984 2.412 0.649 (Source: BEYDOUN & STEPHAN 2005)

70

Table 30 below sums up tables 27 – 29 at national level.

Table 30: National totals for biomass in Forest, OWL and Other Land from NFA.

Type

Above-ground Biomass (million tons)

Below-ground biomass (million tons)

Total living biomass (million tons)

Forest 2.793 0.797 3.59 OWL 0.325 0.082 0.407 OL 2.412 0.649 3.061

(Source: BEYDOUN & STEPHAN 2005) 10.3.3 Above ground woody biomass in Forest and TOF by ecological zone This data was not extracted in the statistical analysis 10.3.4 Precision of biomass estimates In light that this information was virtually lacking prior to TCP/LEB/2903, the NFA represents an improvement in knowledge based on numerous initial assumptions concerning form factors, expansion factors, Wood Densities, root/shoot ratios etc., and with the clear limitations that a 0.04% sample has. These findings should be verified through follow up studies and analysis that would also allow determining the forest type specific average biomass and specifying them further according to GEZ and altitude. A realistic balance between the need for new information and available resources should be struck. Comments on forest species: Biomass in forest areas is influenced by several factors: the abundance of the species and its capacity to produce biomass within a period of time, as well as the extent of exploitation, occurrences of fire, rate of decomposition etc. To a large extent the factors are controlled by forest dynamics and forest management/exploitation. For instance, Quercus calliprinos, although it is the most commonly measured tree in forests, has a lower biomass than Quercus infectoria which is less frequent, but fast growing. The latter is the most common species in deciduous broadleaved forests and has a lower biomass than Quercus cerris. This could be attributed to the fact that Quercus cerris growing stock is overestimated because the measured single stem trees were found mainly in old well-managed stands. Q. Calliprinos and Q. infectoria are found mainly in coppices that are periodically cleared for charcoal and fuel wood production (every 15-25 years). Coniferous trees are of bigger size, and are often rather old (> 70 years), except for Pinus brutia which is exposed to frequent forest fires. Thus cedar and juniper stands have a significant biomass although their areas are small. Pines are fast growing, and cover the largest area among coniferous trees, which makes them the main contributor to biomass of forest species. Further studies should focus on reviewing the different factors and formulas used for the calculation of volume, biomass and carbon stock in order to get accurate data and there is also a need for studies of increment. These studies should concern the ten species cited above as well as Abies cilicica (which was not found as dominant species in any tract, but is an important species in North Lebanon). Those species have the best potential for wood production in forest areas (BEYDOUN &

STEPHAN 2005).

71

Comments on OWL species: The most important species in biomass growing stock in OWL is Juniper trees, which accounts for almost half of the stock. The large trunk that characterize this tree found in higher altitudes shows its importance, and the need to encourage its plantation to preserve the stock and ensure the demand of fuelwood in those cooler zones. Populus alba, which is much less frequent, produces a significant biomass compared to the other species. The plantation of this species in OL is important; especially because it is fast growing. Comments on OL species: Biomass of OL is composed mainly by the biomass provided by fruit trees (olive, citrus, etc) and by trees found in woodlots and in urban areas (Pinus pinea, Quercus calliprinos, Quercus infectoria). Since trees grown as wind breaks, or along roads or rivers, have good growth conditions, they provide an important share of the biomass growing stock (Eucalyptus spp., Casuarina equisetifolia, Cupressus sempervirens, Platanus orientalis, etc.) (BEYDOUN &

STEPHAN 2005). The concerns related to estimating volume in fruit trees mentioned in section 10.2.2 also apply to the estimations of biomass, as the Growing Stock is a component in the equation leading to the estimated biomass. 10.4 Carbon Stock 10.4.1 Carbon Stock by Global Classes The calculation of carbon stock is based on the biomass data. It is calculated by multiplying respectively the above-ground biomass and the below ground biomass with the default value for carbon content in living biomass (50% = 0.5 ton carbon per ton dry weight (Biomass)).

Table 31: National totals for biomass in Forest, OWL and Other Land from NFA.

Type

Carbon of Above-ground Biomass (million tons)

Carbon of Below-ground biomass (million tons)

Total (million tons)

Forest 1.3965 0.3985 1.795 OWL 0.1625 0.041 0.2035 OL 1.206 0.3245 1.5305

(Source: BEYDOUN & STEPHAN 2005) National stock of biomass and carbon of the Lebanese forests and OWL is rather low. This is related to the relatively degraded/unmanaged state of the forest resource, the fact that many natural forest types in Lebanon are open and therefore moderately stocked and due to a modest forest area. The class Other Land contains Fruit orchards which make up a large part of the total growing biomass and carbon stock, which to a certain extent alleviates the pressure on the forest resources. 10.4.2 Precision of Carbon storage estimates The comments of section 10.2.2 and section 10.3.4 apply.

72

10.5 Dbh. Distribution Tree diameter is measured over bark in whole centimetres, at 1.3 meters above the ground (Diameter at breast height or Dbh). The trees in forest and outside forest are divided into 9 diameter classes of 5 cm intervals. Table 32 shows the percentage of trees in each distribution of the diameter classes in forest and trees outside forest.

Table 32: Diameter distribution of measured trees in Forest and TOF

DBH Distribution % of trees measured in forest

(total= 4762)

% of trees measured in TOF

(total= 12228) [10 -15] 42 53.4 [15 - 20] 16 22.4 [20 - 25] 11.6 10.6 [25 – 30] 8.5 5.4 [30 - 35] 8 3 [35 – 40] 6.4 1.7 [40 - 45] 3.6 1.2 [45 – 50] 1.8 0.7 [50 +] 2.1 1.6

(Source: BEYDOUN & STEPHAN 2005)

0

10

20

30

40

50

60

10-15

15-20

20-25

25-30

30-35

35-40

40-45

45–50

50 +

Diameter class

Per

cent

age

of tr

ees

mea

sure

d

Trees in Forest Trees Outside Forests

A distribution with many small/young individuals and fewer large/old individuals is normal for most populations.

73

For forest, the high frequency of trees with small diameters is a result of many comparatively young stands where few individual develop to large sizes. Most of these trees are found in broadleaved coppices that are cut periodically and in regeneration of burnt Pinus brutia stands. The relative low diameter for conifers is related to the absence of silvicultural practices (thinning and selective felling) since coniferous forest exploitation is still prohibited. Trees with large diameters in forests, are mainly those who are found in protected areas or in old stands (cedar, juniper, oak). The same pattern is seen for TOF, which is because most TOF are orchard trees. Orhard trees are pruned below 1.3 meters and felled and replaced when they are no longer productive. TOF therefore show an even higher percentage of small trees and a lower frequence of large trees. Trees outside forest between 25 – 35 cm Dbh, are mainly ornamental trees in urban areas, or trees found in woodlots or along rivers. Outside forests, big trees are mainly old junipers in grasslands with trees, big riparian trees along rivers or old olive trees (BEYDOUN & STEPHAN 2005). 10.6 Biodiversity 10.6.1 List and frequency of tree species in forest by forest type

Biodiversity in Broadleaved Forest Lebanon has specific forest types determined by exposition, altitude and proximity to the coast. The natural vegetation of Lebanon includes a wide range of broadleaved species. The most common inventoried broadleaved trees were Quercus species as shown in table 33:

74

Table 33: Inventoried broadleaved species and their relative frequencies. Scientific name

% frequency of species

( Total Number of trees = 2389)

Endemic species

Quercus calliprinos 40.94

Quercus infectoria 31.6 Quercus cerris (undetermined local subspecies) & var.pseudocerris 14.06 X

Ostrya Carpinifolia 4.27

Prunus ursine 1.381

Quercus brantii var. Look 1.298 X

Platanus orientalis 1.214

Arbutus andrachne 1.172

Pyrus syriaca & var. boveri 1.046 X

Ceratonia siliqua 0.628

Cercis siliquastrum 0.46

Acer syriaca 0.419

Ficus carica 0.293

Prunus amygdalis agrestis 0.251 X

Prunus amygdalis communis 0.251

Styrax officinalis 0.209

Quercus cerris X infectoria 0.084

Rhamnus alaternus 0.084

Celtis australis 0.084

Acer tauricolum 0.084

Malus trilobata 0.042 X

Sorbus torminalis var. pinnatifida 0.042 X

Alnus orientalis 0.042

Populus nigra 0.042 (Source: BEYDOUN & STEPHAN 2005)

According to the list of endemic species in Lebanon, two species were found in the broadleaved forest area, Malus trilobata and Prunus amygdalis agrestis. No species belonging to the red list of the country were found. Quercus calliprinos is the most common species in broadleaved areas. It thrives in all ecological zones up to 1500 m. Like most broadleaved species it subsists mainly as a coppice. It is considered the most common species in Lebanon. Quercus infectoria follows in its relative frequency. It is found between 200 and 1700 m, mostly on relatively deep soils. It is less frequent because its area has been decreasing due to land reclamation, since it is found on more fertile soils.

75

Quercus cerris with its endemic local subspecies comes in the third place, although it is usually a single stem tree, because one tract includes the most important pure stand of this species in North Lebanon. Ostrya carpinifolia is characteristic of the humid areas of subtropical mountains is found in the northern part of the western slopes of Mount Lebanon. It is overestimated by the FRA, because one tract includes one of its important coppice stands in Mount Lebanon. Quercus brantii spp. Look which is an endemic tree found in the southern part of Mount Lebanon and Mount Hermon above 1300 m, is well represented. Another endemic species is Malus trilobata, which is found sporadically in mixed and deciduous forests of the subtropical mountain zone. The remaining species are less frequent either because they are localized in one ecological zone (e.g. Sorbus spp., Acer spp., Ceratonia siliqua) or because they require special edaphic conditions (e.g. Platanus orientalis, Populus spp., Alnus orientalis). Many species, although they thrive in more than one ecological zone, are modestly represented. This could be attributed to the fact that most of them are found in young coppices, with few trees reaching 10cm Dbh (e.g. Rhamnus alaternus, Styrax officinalis, Cercis siliquatrum.). Some species occasionally seen during the field work were not measured since they did not reach 10 cm Dbh or were found outside the plots (e.g. Fraxinus ornus, Sorbus flabellifolia, Rhamnus libanotica, Ribes orientalis, Prunus mahaleb). Many rare species mentioned by some references were not found or maybe not identified within the studied tracts (e.g. Quercus ithaburensis, Quercus pinnatifida, Quercus cedrorum, Quercus libani, Cornus australis, Rhamnus catharica, Amelachier ovalis). Endemism in tree species is only represented by Malus trilobata and Amygdalis agrestis. At subspecies level Quercus brantii ssp. Look, Pyrus syriaca var. Boveii, Sorbus torminalis var. pinnatifida, and an undetermined subspecies of Quercus cerris are represented. In general, trees thriving on the edge of their natural geographical area have different morphological traits such as smaller leaves, fruits and flowers. This leads in certain species to a differentiation classified as subspecies or cultivar. This is the case of species growing normally in humid ecological zones. The number of species trees per hectare does not show mathematically the frequency of occurrence of the species because their respective area of expansion is very fragmented and limited. Stands do not contain all the species. They are very heterogeneous and have limited areas, which makes the extrapolation of the results to hectares unrealistic. For instance, Quercus calliprinos which is a very common tree is mathematically less frequent than its accompanying species Styrax officinalis, which is rare (BEYDOUN & STEPHAN 2005).

76

Biodiversity in Coniferous Forest

Table 34: Inventoried coniferous species and their relative frequencies.

Scientific name % frequency of

species (Total number

of trees = 2344)

Endemic species

Pinus brutia 43.77

Pinus pinea 35.84

Juniperus excelsa 9.73

Cedrus libani 4.44

Juniperus drupacea 3.71

Cupressus sempervirens pyramidalis 1.62

Juniperus oxycedrus & var. refuscens 0.68 X

Cupressus sempervirens horizontalis 0.21 (Source: BEYDOUN & STEPHAN 2005)

No species registered in the coniferous forest area belonged to the red list or to the endemic list of the country. A subspecies could be endemic: Juniperus oxycedrus ssp. refuscens. The relative frequency of tree species and their respective areas are correlated, except for Juniper species since they are not found in dense populations.

Pinus brutia is the most frequent species, in coniferous and mixed forests, due to forest fires and to its plasticity of growth in all ecological zones. It is followed by Pinus pinea which is widely planted for production of pine nuts.

Abies cilicica and Juniperus foetidissima have their southern geographical area limits in Lebanon. A. cilicica is the dominant species in a few stands in North Lebanon, and Juniperus foetidissima is one of its accompanying species. Unfortunately, those stands are outside the studied tracts. Abies cilicica is also found as an accompanying species to the cedar and juniper in North Lebanon.

Cupressuss sempervirens is rarely found in pure stands, but mostly mixed with Pinus brutia in some parts of Lebanon.

Pinus halepensis is found sporadically in South Lebanon, and is not found inside the studied tracts (BEYDOUN & STEPHAN 2005). 10.6.2 List and frequency of tree species outside forest (TOF) Trees Outside Forests are mainly dominated by olives and orchard species as shown in table 35:

77

Table 35: Inventoried TOF species and relative frequencies (BEYDOUN & STEPHAN 2005).

Scientific name

% frequency of species

(Total number of trees=12250)

Endemic species

Olea europea 39.49

Citrus sp 15.87 Quercus calliprinos 5.37

Malus communis/Malus domestica 5.13

Prunus armeniaca 5.01

Juniperus excelsa 2.83

Prunus amygdalis agrestis 2.18 X

Ficus carica 1.95

Cupressus sempervirens 1.73

Quercus infectoria 1.67

Pinus pinea 1.65

Prunus amygdalis communis 1.36

Diospyros kaki 1.32

Prunus ursina 1.30

Pyrus syriaca & var. boveri 1.25

Pinus brutia 1.18

Prunus avium 0.87

Casuarina equisetifolia 0.80

Eucalyptus spp. 0.79

Juglans regia 0.76

Morus alba/Morus nigra 0.70

Platanus orientalis 0.69 Pistacia terebintus ssp. palaestina 0.68

Malus trilobata 0.58 X

Prunus amygdalis 0.58

Persea americana/Persea gratissima 0.42

Pyrus communis 0.35

Melia azedarach 0.33

Ceratonia siliqua 0.24

Juniperus oxycedrus 0.24

Cupressus sempervirens pyramidalis 0.21

Populus tremulla 0.21

Populus nigra 0.20

Acer syriaca 0.16

Alnus orientalis 0.14

Populus alba/bolleana 0.13

78

Errobotrium japonica 0.12

Tilia spp 0.11

Crateagus monogyna 0.11

Prunus domestica 0.11

Prunus amygdalis korschinskii & orientalis

0.10

Accacia mimosa/albida/farnesiana 0.07

Juniperus drupacea 0.07

Prunus cerasifora 0.07

Pinus halepensis 0.07

Punica granatum 0.07

Laurus nobilis 0.06

Pinus canariensis 0.06

Prunus persica 0.06

Ziziphus jujuba 0.06

Prunus cerasia 0.05 (X)

Quercus brantii var. Look 0.05 X

Eleagnus angustifolia 0.04

Quercus cerris (local ssp) & var.pseudocerris

0.04 X

Castanea sativa 0.04

"Not known" 0.03

Celtis australis 0.02

Cercis siliquastrum 0.02

Rhus coriaria 0.02

Cedrus libani 0.02

Rhamnus alaternus 0.02

Robinia pseudoaccacia 0.02

Salix acmophylla 0.02

Vitis sp 0.02

Arbutus andrachne 0.01

Cornus australis 0.01

Crataegus azerolus & sinaica 0.01

Jacaranda mimosifolia 0.01

Juniperus foetidissima 0.01

Myrtus communis 0.01

Phillyrea media 0.01

Salix australior & dinsmorei 0.01

Salix libani 0.01

Styrax officinalis 0.01

Ulmus minor 0.01

79

TOF species are mainly fruit species in agricultural lands with Olive, citrus, apple, apricot, almond and fig being the most frequent. Forest trees are found in OWL and fragmented woodlots, frequent in the surroundings of urban areas (e.g. Quercus calliprinos, Quercus infectoria, Pinus pinea.). Evergreen oak, stone pine, laurel, nettle tree and cypress are also frequent in urban areas, and as isolated trees planted near churches, cemeteries, roadsides, monasteries and houses. Many species are mainly found as accompanying species to oak in OWL (e.g. Arbutus andrachne, Rhus coriaria, Phillerea media, Rhamnus alaternus, Myrtus communis, Styrax officinalis, Crataegus spp., Acer syriaca, Laurus nobilis.) and rarely reach 10 cm Dbh in such formations, which explains their relatively low frequency.

Other species are more frequent in grasslands with trees (e.g. Juniperus spp, Prunus ursina, Pyrus syriaca, Ceratonia siliqua, Malus trilobata, Amygdalus spp., Crataegus spp.). This could be explained by the residues of the agri-silvo-pastoral system that preserved wild fruit trees near agricultural and pastoral lands.

Riparian species are found along rivers or near water springs (e.g. Populus spp., Platanus orientalis, Salix spp., Eleagnus angustifolia, Alnus orientalis).

Introduced species are few, among them Tilia spp., Eucalyptus spp., Acacia spp, and Casuarina equisetifolia. The latter is planted as a wind breaking curtains while the others along roadsides. Others are planted as ornamentals in house yards like Jacaranda mimosifolia. Few coniferous species have been introduced for reforestation purposes, among them Pinus canariensis.

Endemism is rare at species level, but is more frequent in sub-species or cultivars. Among species we find Prunus cerasia (MOUTERDE 1966 – 80). It is a local plum tree that has a non-identified origin and that has become very rare in the orchards. As for sub-species and cultivars we mention Populus bolleana, Salix alba var. micans or libanotica, Pyrus syriaca var. bovei (in Anti Lebanon), and Prunus ursina forma leioclada, all of them are generally found outside forests. The areas of expansion are very fragmented and limited (BEYDOUN & STEPHAN 2005). The NFA focuses on the woody vegetation and the relative occurrence of Forest and TOF species – for which it provides an initial first rough overview. A more complete picture of the biodiversity should include additional aspects of the occurrence of flora, fauna and fungi – but lies beyond the scope of the current NFA.

80

10.6.3 Forest by stand structure option

Figure 12: Stand structure by relative area of forest type (Level 2)

Stand Structure

BroadleavedConiferousMixed Forest

Rel

ativ

e Fo

rest

Are

a

100

80

60

40

20

0

Single layer

two-layer vegetation

13

50

38

87

50

63

(Source: BEYDOUN & STEPHAN 2005) The frequency of two layers of trees is rare in Lebanon. In most cases, the lower layer is made of shrubs. Two layers vegetation is frequent in Pinus pinea stands where trees are pruned thus allowing sun rays to reach the ground, which enables other species to thrive underneath. Pinus pinea is planted and most of the lower layer is composed of broadleaved species. When pine forest is not maintained, the lower layer forms a second layer of trees and woody vegetation. The frequency of fire in mixed forests of Pinus brutia and broadleaved species in the subtropical dry forest enhances the regeneration of the fast growing Pinus brutia, and degrades the broadleaved tree species, that will only create a lower layer of shrubs. If the canopy cover of Pinus brutia trees is not dense, and the broadleaved trees layer is more than 25%, they are mixed forests, but the team leaders would also interpret them as a second lower layer during fieldwork. This explains also the high percentage of double layers in mixed forests. In coniferous dense stands and evergreen broadleaved stands, there is only one single layer. A second layer may occur in open coniferous stands and deciduous broadleaved stands, where the light intensity allows a lower layer of different trees, shrubs and ferns to develop (e.g. Quercus, spp., Rosa spp, Myrtus spp, Pistacia spp, Styrax officinalis, Lonicera spp, Cistus spp.). Stands with two layers exist only in the dry subtropical and subtropical mountain ecological zones. Thus, on the eastern slopes of Mount Lebanon, the Bekaa and Anti-Lebanon, only forests with single stands exist (BEYDOUN & STEPHAN 2005).

81

10.6.4 Forest by human induced disturbance level. Disturbance is defined as the impact level of the human activities in the forest. The analysis of collected data shows that 85.1 % of the forest area (app. 118,609 hectares) is disturbed while 14.9 % (20,797 ha) are classified as undisturbed. This disturbed area is divided into three categories: Slightly disturbed, moderately and heavily disturbed as shown in table 36.

Table 36: Forest by disturbance level

Forest disturbances Size x % Disturbed forest area

Slightly disturbed 53 41.2 48,867 Moderately disturbed 53 55.6 65,947 Heavily disturbed 53 3.2 3,795 Total 100 118,609

(Source: BEYDOUN & STEPHAN 2005)

Table 37: Forest type (Level 2) by disturbance level

Coniferous forest

Broadleaved forest Mixed forest Forest disturbances

x % (Ha) x % (Ha) x % (Ha) No disturbances of forest 24.7 11,085 23.5 18,538 3.8 593 Slightly disturbed 37.1 16,650 41.2 32,501 34.5 5,385 Moderately disturbed 58.4 26,209 57.8 45,597 53.6 8,367 Heavily disturbed 4.5 2,020 1.0 789 11.9 1,858 Total 100 44,879 100 78,887 100 15,610

(Source: BEYDOUN & STEPHAN 2005) Undisturbed forests are areas that do not show any damage or symptom of human activities. In general, such forests are those which are not used for production purposes and they are limited to inaccessible and protected areas. Disturbed forests include all forests having a production status. Since forest utilization is normally not associated with any defined management that ensures the sustainability of the natural resources and the land use, those forests are considered slightly or moderately disturbed in the field survey. This explains the even distribution of percentages between the three types of forests. Heavily disturbed forests are areas showing severe degradation in the tree health, or the accompanying vegetation (fires, insect outbreaks, landfills, quarries, water and soil pollution, intensive urban activities, etc). The susceptibility of coniferous trees to such influences and the encroachment of mixed forests by urban areas explain the higher percentage of heavily disturbances in coniferous and mixed forests. Another reason is that Pinus pinea and Pinus brutia are not a typical vegetation formation of the Mediterranean coastal zone of Lebanon, which is more represented by the evergreen broadleaved coppices. The first is widely planted by man for its nuts production, and the latter spreads faster from the degraded coppice after induced fires (BEYDOUN & STEPHAN 2005).

82

The results have been extrapolated from a 0.04% sample which naturally cause some imprecision. The fact that a very large percentage of the Lebanese forests are categorized as disturbed is not surprising considering the ancient and at times turbulent history of settlement in the country. 10.7 Use of resources Two main categories are used to classify products: Wood and Non-Wood products. Information on areas accessible for wood supply is important for land use planning, for development of sustainable forest utilization and for a policy perspective. Non-wood forest products (NWFPs) play an important role in the daily life as sources of food and income, especially in rural areas. The relative areas of exploitation of each product is expressed relative to the area of the Land Use Class and the total country area (1,045,200 ha) in each land use class. Table 38: Relative areas of exploitation of Wood and Non Wood Forest Products

Wood and Non wood products

% Land Use Area % country area

Forest OWL OL Fuel wood 38 26 4 10.84 Plant food 11 17 36.3 30.91 Honey, wax 5.7 6 1 2.13 Plant medicines 3 0 0.5 .80 Timber 2 0 0.3 .50 Charcoal 2 0 0 .27 Ornamentals 1.2 0 2.4 1.96 Fodder 1 3.5 14 11.17 Bush meat 0.7 1 0 .20 Soap 0.2 0 1.5 1.16

(Source: BEYDOUN & STEPHAN 2005) Plant food is the most common group of Products/Services as it is exploited/used in 31% of the total country area. Fuel wood is the product most commonly provided by the forest area.

Plant food products occupy the largest area of the country, which is mainly found in the agricultural areas (27% of the total surface of the country) and Pinus pinea forests (more than 7000 ha). It is followed by fodder products, also cultivated or found in grasslands and some OWL. Fuel wood occupies the third place in surface area and is extracted from Forests, OWL and OL. Such distribution is related to the prohibition of forest exploitation and the high rate of orchard renovation which leads to the use of fruit tree wood as fuel wood. The remaining products and services have limited areas of extraction because of the limited area of Forests and OWL (13% and 10% respectively). (BEYDOUN & STEPHAN 2005).

83

10.7.1 Wood and NWFP provided by the forest The rank of products according to their percentage of exploitation is compiled based on the results of interviews of key informants categorized as follows: products of high, medium and low importance.

Table 39: Exploitation of Wood and Non-wood forest products (ranked) presented as relative forest areas in which the products are exploited

Wood and Non wood products

% by forest Area

% Area Rank

Total High Medium Low Fuel wood 38 23 13 2 Plant food 11 5 6 0 Honey, wax 5.7 1.2 3.3 1.2 Plant medicines 3 0 2 1 Timber 2 0 2 0 Charcoal 2 2 0 0 Ornamentals 1.2 1.2 0 0 Fodder 1 0 1 0 Bush meat 0.7 0 0.7 0 Soap and cosmetics 0.2 0 0.2 0

(Source: BEYDOUN & STEPHAN 2005) 10.7.2 Wood and NWFP provided by the TOF

Table 40:Exploitation of Wood and Non-wood forest products (ranked) presented as relative Other Wooded Land areas in which the products are exploited

Wood and Non wood products

% by OWL Area

% Area Rank

Total High Medium Low Fuel wood 26 18 6 2 Plant food 17 6 10 1 Honey, wax 6 1 2 3 Plant medicines 0 0 0 0 Timber 0 0 0 0 Charcoal 0 0 0 0 Ornamentals 0 0 0 0 Fodder 3.5 3 0 0.5 Bush meat 1 0 1 0 Soap and cosmetics 0 0 0 0

(Source: BEYDOUN & STEPHAN 2005)

84

Table 41: Exploitation of Wood and Non-wood forest products (ranked) presented as relative Other Land areas in which the products are exploited

Wood and Non wood products

% by OL Area

% Area Rank

Total High Medium Low Fuel wood 4 1.5 2.5 0 Plant food 36.3 34 2 0.3 Honey, wax 1 0.3 0.7 0 Plant medicines 0.5 0.4 0 0.1 Timber 0.3 0.3 0 0 Charcoal 0 0 0 0 Ornamentals 2.4 1.5 0.5 0.4 Fodder 14 12.5 1 1.5 Bush meat 0 0 0 0 Soap and cosmetics 1.5 1 0.5 0

(Source: BEYDOUN & STEPHAN 2005) Fuel wood is the major product in forests and OWL. Timber and charcoal production were prohibited at the time of the fieldwork and the modest production that was found was considered an illegal production. The legal amendments that have occurred since September 2004 allow charcoal production in broadleaved forests. A significant amount of fuelwood is extracted from felled and pruned trees found in OL (orchards). Plant food is the major product that is taken from OL and ranks second in forests and OWL. In OL plant food is provided mainly by agricultural zones, in forests by pine and carob trees, while in OWL it is provided by carob trees, laurel, sumac and wild plants. This product ranks high in forests, because Pinus pinea forests are planted in the purpose of pine nut production. Honey production is significant in Lebanon, and widely distributed in all land uses. The number of bee hives is around 120000 all over the country which makes an average of 12 hives/km2 or 12 hives per tract, i.e. a significant frequency of bee hives. Plant medicines are more frequently extracted from forest areas, rather than OWL or OL. The major product is sage which cutting is regulated by the State. Other species are extracted and used locally for infusions (mainly Eleagnus angutifolia and Crataegus spp., from tree species and an undetermined number of wild plants). Many species found in grassland in high altitudes have medicinal properties and many of them are endemic (i.e. Ferula hermonensis). The values of such products are still underestimated in Lebanon. Also many species are cultivated for aromatic and medicinal use (Rosa damascena and sentifolia, Sambucus nigra, Laurus nobilis etc). Fodder extraction is more frequent in OWL than in forest, because the shepherds have easier access to these areas. Fodder in OL is mostly cultivated or extracted from grasslands. Ornamentals are planted in OL where their extraction is the most frequent. The wild vegetation in forests and OWL seems to be of lower importance (ferns, flowers, shoots of Myrtus communis, Ceratonia siliqua, etc).

85

Bush meat (e.g. wild boar, birds, rabbit) and products used for soap and cosmetics industries are almost negligible importance in Lebanon, except for soap production from olive oil (OL zones) (BEYDOUN & STEPHAN 2005). 10.7.3 Social, economic and environmental services provided by the forest The relative areas of exploitation of each service is expressed relative to the area of the Land Use Class and the total country area (1,045,200 ha) in each land use class.

Table 42: Relative areas of exploitation of Services

Services % Land Use Area

% Country area

Forest OWL OL Soil and water conservation 47 30 4 12.43 Grazing 20 40 19 21.31 Recreation and tourism 4 0 1 1.30 Hunting (sport) 2 6 1 1.65 Source for employment (paid) 2 0 0.2 0.42 Windbreaks 1 0 1 0.90 Conservation 1 0 0 0.13 Shade 0 1 5 3.92

(Source: BEYDOUN & STEPHAN 2005)

Table 43: Exploitation of Services (ranked) presented as relative Forest areas in which the Services are used

Services % by forest Area

% Area Rank

Total High Medium Low Soil and water conservation 47 29 9 9 Grazing 20 13 5 2 Recreation and tourism 4 1 2 1 Hunting (sport) 2 1 0 1 Source for employment (paid) 2 0 1 1 Windbreaks 1 0 1 0 Conservation 1 1 0 0

(Source: BEYDOUN & STEPHAN 2005)

86

10.7.4 Social, economic and environmental services provided by TOF

Table 44: Exploitation of Services (ranked) presented as relative Other Wooded Land areas in which the Services are used

Services type % by OWL

Area % Area Rank

Total High Medium Low Grazing 40 30 9 1 Soil and water conservation 30 17 9 4 Hunting (sport) 6 2 1 3 Shade 1 0 1 0 Windbreaks 0 0 0 0 Source for employment (paid) 0 0 0 0 Conservation 0 0 0 0

(Source: BEYDOUN & STEPHAN 2005)

Table 45: Exploitation of Services (ranked) presented as relative Other Land areas in which the Services are used

Services type % by OL Area

% Area Rank

Total High Medium Low Grazing 19 18 1 0 Shade 5 1.6 3 0.4 Soil and water conservation 4 1 2 1 Windbreaks 1 0.8 0.2 0 Hunting (sport) 1 0.5 0 0.5 Recreation and tourism 1 0.5 0 0.5 Source for employment (paid) 0.2 0 0 0.2 Conservation 0 0 0 0

(Source: BEYDOUN & STEPHAN 2005) Soil and water conservation is the most important service provided by the forest in Lebanon. It is ranked in the second place in OWL. This due to the fact that most forests are situated on steep slopes, where soil and water conservation is a priority, and where other services cannot compete due to difficult access. It also ranks third in OL in woodlots, grassland with trees, or in some urban areas where trees are present. Grazing shows an opposite picture; it is first ranked in OL, mainly grasslands and some agricultural lands, which constitute the main areas of OL. In OWL, grazing also ranks first. The open nature of the woods makes grazing easier and fodder species are more frequent, and at the same time, the presence of grazing animals (especially goats) reduces tree regeneration and tree height which degrades the forest into OWL. In forest areas, soil and water conservation is followed in ranking by grazing. This activity is more frequent in open non-protected forests where animals can find fodder plants to graze. The big difference between “soil and water conservation” and “grazing” means that the latter service is becoming more and more restricted in forest areas due to the protection of forests against exploitation and grazing.

87

Recreation and tourism ranks third inside forests which illustrates the growing importance of such activities in Lebanon. Hunting is a prohibited activity, it is however reported present, probably due to the absence of a clear law for hunting. Inside the forest and in OWL most hunters hunt bush meat (e.g. wild boar, rabbits and birds). Source of employment in the form of e.g. pine nut collectors, charcoal producers, guides and forest guards, handicraft workers were not ranked in the first three places, and in most cases, the people live outside of the tracts/plots (BEYDOUN & STEPHAN 2005). 10.7.5 Users of forests and tree resources Products: User rights are in general exclusive to land owner. For detailed analysis of user rights in Forest, OWL and OL please refer to (BEYDOUN & STEPHAN 2005). Men do the major part of the exploitation of wood and non-wood forest products. Women and children participate in OWL and mostly OL where the products are less heavy to extract (fruit and olive collection, olive and laurel soap fabrication, wild plants collection, ornamentals collection). The enterprise form is mostly categorized as spontaneous for product extraction inside forests and OWL. In some cases where pine nuts or fuel wood is managed by municipalities or private companies extracting carob molasses, sage, honey, wax, etc. it can be organized. In OL, the percentage of organized forms of extraction is higher since all agriculture products (soap and cosmetics, food, fodder, ornamentals) are considered organized enterprises. Services Activities such as source of employment, grazing, wind breaks, conservation, recreation and tourism and shade have an exclusive right for the land owner and a common right in communal and State lands. Most small landowners allow activities such as grazing, recreation and tourism on their land, because the relatively small size of their land which does not allow them to get a significant income by renting it to other users. Such activities can be forbidden if the land owner considers them to cause harm to his land and/or wellbeing. The exclusive right of grazing, recreation and tourism occurs mainly when such activities are organized, and in big land parcels (i.e. farms, communal lands, skiing stations, etc). Hunting is prohibited by law in Lebanon – however it occurs as a spontaneous activity. Seasonal labour is needed to maintain TOF and harvest their products (e.g. planting, pruning, irrigating, spraying, harvesting, etc). Forests provide a source of employment in protected areas (forests guides, guards, scientists, etc), and for wood and non-wood products exploitation (charcoal, timber, pine nuts, sage etc.), the data collected during the survey however do not reflect these employment opportunities.

88

Grazing occurs mainly spontaneously in Forests and OWL and while in OL it occurs also as an organized enterprise, (where modern farms exist and in pastures owned by municipalities where grazing is an organized activity by the municipality). Men function as shepherds to watch the grazing animals, collect fodder and milk them. Women are responsible for milk processing. Children participate sometimes in this family business (BEYDOUN & STEPHAN 2005). 10.7.6 Reliability of estimates of forest use attributes The figures from the fieldwork should be seen as indicators only. The figures from the NFA point out the main products and services of Forest and TOF. The percentage of land use area and the percentage of the country area show the frequency of the occurrence of the products and services (P/S) as reported by the NFA. During fieldwork, the team leaders did not rank all P/S, but estimated the 3-5 most important P/S and filled them in succession according to their importance, which may have caused minor products and services not to be included. The interview techniques and possibly reluctance of the interviewed informant to inform on illegal activities (e.g. wood extraction and hunting) and conflicts may well result in a general underestimation of the P/S as well as existing conflicts related to illegal activities in the various land uses. More detailed studies are needed to obtain figures of use for management purposes.

89

11. Forest Fires The area of forest affected by fires and the presence of fires in the forest type was estimated during the fieldwork. The results obtained are shown in table 46:

Table 46: Area estimates of fire occurrences by forest type (level 2)

Forest % Ha

Coniferous forest

% Ha

Broadleaved forest

% Ha

Mixed forest

% Ha No evidence of fire 85.9 119,72

4 25.9 36,098 52 72,476 8 11,150

Recent fire 0 0 0 0 0 0 0 0 Old fire 14.1 19,652 14.097 19,648 0 0 0.003 4

(Source: BEYDOUN & STEPHAN 2005)

The percentage of the forest area affected by fire is about 14.1 % (or 19,652 hectares). This fire area is divided into 14.097% of coniferous fire area estimated (or 19,648 hectares); the rest of this fire forest area is from 0.003% mixed forest (or 4 hectares). Forest fires are almost exclusively limited to coniferous stands - below typical fires in pine stands are mentioned. A common factor is that uncontrolled forest fires almost always develop from fires started by humans. Uncontrolled forest fires in Pinus pinea stands are started by farmers to eliminate the lower layers of vegetation, since it is more cost-effective than pruning or cutting. Charcoal production was prohibited until September 2004, which caused the farmers to burn the broadleaved layer instead of cutting it. Uncontrolled forest fires in Pinus brutia stands are mostly unintentional, and occur in areas of encroachment of pine forests with urban and agricultural areas in Mount Lebanon and North Lebanon. Again, the absence of any economical interest related to the management of forests leads to the absence of pruning and clearing, which exposes the forest to an increasing risk of fire. No evidence of fire was registered in the broadleaved forest area and no evidence of recent fires was encountered. The absence of recent fires is attributed to the season of field work (November to July) while forest fire season in Lebanon starts in August and ends with the first rains of autumn (BEYDOUN & STEPHAN 2005).

90

12. Local Population 12.1 Trends Major historical events affecting the local people and the settlement history / time when the land was occupied was noted where possible during the fieldwork. The frequency of the different settlement history is shown below in table 47.

Table 47 Relative distribution of settlement history from NFA for the populated areas. Settlement History Percentage

Land Tenure Change 1.8

Wars/Armed conflicts 9.2

Socio-economical crisis 11

Agricultural contraction 20

Urban expansion 21.1

Agricultural expansions 36.7

Total 100 (Source: BEYDOUN & STEPHAN 2005) According to the NFA, agricultural expansion has been the cause of most settlement (36.7%) in the sampled tracts. In reality the cause of settlement changes can be more than one factor alone and they are often interrelated. The history of settlements in Lebanon is as old as the Neolithic era. Some cities figure among the oldest in the World (e.g. Byblos). In general, settlements occurred first on the coastal area, then in the Bekaa valley. The mountains were the place of hunters, carpenters and shepherds until the 4th century A.D. Mount Lebanon became the shelter of different communities escaping from persecution. Between the 17th and the 19th century, agricultural expansion throughout the country increased the number of villages. From late 19th century until the end of the First World War, many settlements were abandoned due to conflicts and economical crisis, and rural population migration was very intense. Between 1920 and 1975, agriculture expansion in irrigated zones fixed and developed many settlements in rural areas, while cities expanded their surfaces due to agriculture contraction in rain fed zones. Since 1975 urban expansion increased dramatically in the areas that hosted displaced populations during the war (BEYDOUN & STEPHAN 2005). 12.2 Activities The Main Activity is defined as the main source of income generation and employment of the permanent population living within tract. The most common main activity is agriculture (58%) followed by urban and semi-urban activities (28%). The Overall Main Activity is the main income generation and employment sources of most of the total population including both permanent and seasonal population living within tract.

91

Table 48: Relative distribution of the population main activities Population Main activity Percentage

Tourism 1.8

Forestry 2.8

Live stock 9.2

Urban or semi urban 27.5

Agriculture 57.8

Not Known 0.9

Total 100 (Source: BEYDOUN & STEPHAN 2005)

According to data collected by the interviews conducted during the NFA, 59.6% of the populated area had experienced increasing population over the last five years. 27.5 % was reported as having experienced no change in population and a decrease in population was registered for 8.3 % of the populated area. For 3.7% it was not known whether the population had increased or decreased.

Since the tracts are mostly outside the big cities (Beirut, Tripoli, Saida, Zahleh, etc.), the major activity found within tracts is agriculture. Nevertheless, urban and semi urban activities have an important share which illustrates the extent of urban sprawl in Lebanon. Livestock activities are represented especially in tracts found in high altitudes in Mount Lebanon and Anti-Lebanon, while forestry is much localized to tracts found in Pinus pinea forests (BEYDOUN & STEPHAN 2005). Table 49: Relative distribution of the population dynamics by the permanent population’s main activity (107 tracts)

Population Dynamics Population Main activities

Increasing (%)

Not changing (%)

Decreasing (%)

Not known (%)

Urban 72.4 13.8 13.8 Agriculture 61.9 31.7 4.8 1.6 Livestock 40 40 20 Forestry 33.3 33.3 33.3 Tourism 50 50 (Source: BEYDOUN & STEPHAN 2005)

92

Table 50: Relative distribution of the population dynamics by the total population’s main activity (107 tracts)

Population Dynamics Population Main activities

Increasing (%)

Not changing (%)

Decreasing (%)

Not known (%)

Urban 61.9 21.4 9.5 7.1 Agriculture 61.8 30.9 5.5 1.8 Forestry 50 50 Livestock 37.5 37.5 25 (Source: BEYDOUN & STEPHAN 2005)

The NFA findings concerning population dynamics in both total and permanent population main activities show that the population having urban activities is increasing the most. This is due to population increase and also to the fact that in Lebanon the sector of trade and services consists of more than 70% of the GDP, and attracts most of the manpower and investments. The data collected on population dynamics reflect the migration from rural areas to the cities, emigration and people living in rural areas but working in towns which is pronounced due to small size of country and good infrastructure. Stable population dynamics is found in a large part of the areas where agriculture, livestock, forestry and tourism are the main activities. In such areas most of the rural population growth is equalized by migration and the abandon of such activities towards more profitable ones. Forestry and tourism activities do not show any decrease in the population dynamics which could be related to the sectors being relative minor and the population having an important additional income from other activities (agriculture, urban) (BEYDOUN & STEPHAN 2005).

93

13. Accessibility to tracts and plots The distance from tract centre to different services e.g. hospitals, schools, settlements, markets and roads was registered as measures of tract accessibility. The distances are classified in 2.5 km classes. The histograms show if there is a normal distribution and the averages and the standard deviation of each distance can be found.

13.1 Accessibility to hospitals and schools The accessibility to hospitals is shown in figure 13 together with a curve for the normal distribution. The average distance from the tract centre to the closest hospital is 7.7 km.

Figure 13 Distance to closest hospital (no. of tracts on y-axis)

Distance to closest Hospital (km)

40.0

37.5

35.0

32.5

30.0

27.5

25.0

22.5

20.0

17.5

15.0

12.5

10.0

7.5

5.0

2.5

0.0

70

60

50

40

30

20

10

0

Std. Dev = 6.45

Mean = 7.7

N = 210.00

(Source: BEYDOUN & STEPHAN 2005)

The accessibility to schools is shown in figure 14 together with a curve for the normal distribution. The average distance from the tract centre to the closest school is 3.2 km.

94

Figure 14: Distance from tract centre to closest school (no. of tracts on y-axis).

Distance to closest School (km)

30.027.5

25.022.5

20.017.5

15.012.5

10.07.5

5.02.5

0.0

100

80

60

40

20

0

Std. Dev = 4.50

Mean = 3.2

N = 212.00

(Source: BEYDOUN & STEPHAN 2005)

13.2 Accessibility to Settlements and markets. The accessibility to markets is shown in figure 15 together with a curve for the normal distribution. The average distance from the tract centre to the closest market is 4.2 km.

Figure 15: Distance from tract centre to closest market (no. of tracts on y-axis).

Distance to closest Market (km)

30.027.5

25.022.5

20.017.5

15.012.5

10.07.5

5.02.5

0.0

80

60

40

20

0

Std. Dev = 5.10

Mean = 4.2

N = 211.00

(Source: BEYDOUN & STEPHAN 2005)

95

The accessibility to settlements is shown as staples in figure 16 together with a curve for the normal distribution. The average distance from the tract centre to the closest settlement is 1.7 km.

Figure 16: Distance from tract centre to closest settlement (no. of tracts on y-axis)

Distance to closest Settlement (km)

30.0

27.5

25.0

22.5

20.0

17.5

15.0

12.5

10.0

7.5

5.0

2.5

0.0

160

140

120

100

80

60

40

20

0

Std. Dev = 3.81

Mean = 1.7

N = 211.00

(Source: BEYDOUN & STEPHAN 2005) 13.3 Accessibility to all weather roads and seasonal roads. The accessibility to all-weather roads is shown as staples in figure 17 together with a curve for the normal distribution. Average distance from the tract centre to the closest all-weather road is 0.9 km.

Figure 17: Distance from tract centre to closest all weather road (no. of tracts on y-axis)

Distance to closest All-Weather Road (km)

30.0

27.5

25.0

22.5

20.0

17.5

15.0

12.5

10.0

7.5

5.0

2.5

0.0

200

100

0

Std. Dev = 3.33

Mean = .9

N = 213.00

(Source: BEYDOUN & STEPHAN 2005)

96

The accessibility to seasonal roads is shown in figure 18 together with a curve for the normal distribution. The average distance from the tract centre to the closest seasonal road is 0.5 km.

Figure 18: Distance from tract centre to closest seasonal road (no. of tracts on y-axis)

Distance to closest Seasonal Road (km)

16.515.5

14.513.5

12.511.5

10.59.5

8.57.5

6.55.5

4.53.5

2.51.5

.5

300

200

100

0

Std. Dev = 2.86

Mean = .5

N = 213.00

(Source: BEYDOUN & STEPHAN 2005)

The distances to hospitals, schools, markets, and settlements are relatively short in Lebanon. This is mainly due to the high density of the population, and to the level of urbanization. The average population density exceeds 400 inhabitants/km2, and the percentage of urban population is 88%. These two factors make the road network density very high, and thus the distance from tract centres to all weather roads and seasonal roads has an average of less than 1 km. This facilitated field work conditions as plot localization and access was comparatively easy.

97

14. Comparison of NFA results with existing information The current NFA estimates the forest area of Lebanon at 13.3 % of the total area of the country, while the DFM determines the forest area at 13.1 %. The NFA estimates the OWL area of Lebanon at 10.4 % of the total area of the country, while the DFM determines the forest area at 11.8 %. Former reports and estimates showed figures for forest areas between 5 and 7%. The National classes reported for FRA 2000 was based on the 1965 forest type map and provided the following figures were reported:

Table 51: National Classes reported for FRA 2000 National Class Area (000 ha) Oak forest 40.0 Pines on Limestone (Pinus brutia especially) 5.0 Pines on sandstone (Pinus pinea) 12.0 Cedar forests 2.0 Fir forests (Abies cilicica) 1.0 Juniper forest 9.0 Cypress forest 0.5 Total 69.5

(Source: FAO 2004A) Translated into FRA global classes the forest resources of Lebanon were interpreted as follows for the FRA2000:

Table 52: Global Classes reported for FRA 2000 FRA2000 Global class Area (000 ha) Forest 36 Other Wooded Land 35 Other Land 954 Inland Water 17 FAOSTAT Area 1040

Source: FAO 2001 The Corresponding figures from the NFA and the DFM were:

Table 53: Global Classes from NFA and DFM. FRA Global class NFA

Area (ha) DFM Area (ha)

Forest 139,376 134,372 Other Wooded Land 108,378 120,574 Other Land 797,152 766289 Inland Water 294 1745

NB Area figures are not adjusted to the FAOSTAT area

98

The findings of TCP/LEB/2903 are therefore in sharp contrast to the previous generally quoted datasets for reporting on the forest cover of Lebanon. The figures from both the NFA and the DFM point at a forest resource which area wise is approximately 4 times bigger than that reported for FRA2000 and an area of OWL which is app. 3 times bigger. As both the NFA and the DFM arrive at very similar figures at the Global Class Level it is likely that the forest and TOF resources of Lebanon reported for FRA 2000 were underestimated. The difference between former estimations and the result of the NFA could be attributed to the following (BEYDOUN & STEPHAN 2005): • The fact that a dataset from the 1960s was used for reporting in 2000 - Rapid urbanization with

a considerable rural migration caused abandoning of agricultural lands and pastoral activities followed by invasion of many areas by forest during the second part of the 20th century.

• Differences in definitions for “forest” and “other wooded lands” for Lebanon in earlier estimates and possibly also unclear definitions of OWL.

• Reforestation activities and relative protection of many coniferous stands from cutting and felling during the past decades.

Being based on the spatial database of the Land Use Land Cover Project there is a good compliance between the figures arising from TCP/LEB/2903 and the findings of TCP/LEB/2801. Concerning information related to timber volume, state of the forest and TOF resources and their importance to the local population TCP/LEB/2903 represents the first nationwide systematic sampling of such data. It is therefore not possible to compare with earlier data. The establishment of a system of permanent sampling sites makes it possible to monitor the changes in parameters related to the above – through future repetitions. In general considering the limited availability of well documented information on the Forest and TOF resources of Lebanon prior to TCP/LEB/2903 the project has generated much new information – information that will increase in value with every re-measurement. Concerning biodiversity – several studies have earlier been performed on the diversity of the vegetation of Lebanon e.g. Biodiversity of Lebanon (UNEP & Ministry of Agriculture, 1996). The focus of the NFA has been on the woody biodiversity i.e. on the relative occurrence of woody Forest and TOF species. A more complete picture of the biodiversity would need to include additional aspects of the flora as well as information on fauna and fungi. This lies much beyond the initial scope of the NFA. The established system of permanent sampling sites could function as a starting point for additional studies.

99

15. Capacity building According to the Project Document output 1, mentioned in Chapter 3, one of the major project outputs is to enhance and strengthen the capacity of the Forestry Department of the DRDNR to plan and implement forest inventories, monitor the resources, manage the related information, and contribute to advance sustainable forest and tree management by enabling an increased use of forestry knowledge in forest policy development. As TCP/LEB/2903 is the first time a systematic National Forest and TOF Assessment based on field measurements, interviews and observations was implemented in Lebanon, the starting point was therefore one where the basics of practical fieldwork had to be built up first. The lack of tradition in the field of NFA was however outweighed by the very motivated, well educated staff of the Forest Department of DRDNR and the Regional Development Services of MOA who under the strong leadership of the National Project Coordinator quickly picked up on the various methodologies. The current chapter describes the various capacity building aspects of TCP/LEB/2903. 15.1 Capacity to undertake fieldwork At the start of the project period the tradition for undertaking systematic NFA was non-existent. The field teams were initially trained in the various measurement techniques. The various training aspects of the project are described in section 7.2. The most valuable training however came with experience through the field work supported by the continual supervision of the NC-FI and the TCDC-expert. Slope correction of the measured distances proved somewhat of a challenge for the teams initially as the method used relied on the slope correction table of the field manual. The TCDC-expert introduced break-chaining i.e. measuring slopes in horizontal steps, which for short distances and fragmented landscape is an advantage as there is no subsequent adjustment of distance. The field teams quickly became very good in navigational skills, i.e. using GPS, compass and the georeferenced field maps, although the terrain could pose serious problems and sometimes render parts of plots inaccessible due to slope. At the end of the fieldwork phase, the teams performed well and delivered high quality field data and through the built up experience the field work was also performed more rationally. One major asset throughout the fieldwork phase was that the field teams knew the areas and surroundings well as they were based in the RDS of the Mohafaza. 15.2 Capacity to register and enter data The quality of the submitted reports improved with field experience. Through intensive checking of submitted reports by TCDC-expert and NC-FI the team leaders received instructions on what items were in need of correction with instructions on how to complete them. By the end of the fieldwork phase, the errors contained in the submitted reports were minor. Field forms were only cleared for entry into the database once cleared by both the TCDC-expert and the NC-FI.

100

The team leaders had responsibility for entering the field data in the database application developed by FORM and installed on the Team Leaders computers at the RDS’s. In this connection the team leaders got acquainted with the Access Database, with the importance of data discipline and the importance of entering data as soon as possible after completion of a tract while the work was still fresh in memory. As mentioned in section 9.2 data-entry proved to be a bottleneck for the 2nd project phase. Part of the explanation was the late establishment of functional facilities (PCs and database application at the RDS) for decentralized data entry. Part of the explanation was also that after long days of field work and with a multitude of other responsibilities, it was not always possible for the field team leaders to find the time to enter data after a plot had been completed. It was therefore necessary to recruit additional assistance to assist in the final phases of data entry in order to have all field data entered by late September 2004. Considering that the starting point for some of the field team leaders was zero direct database experience, the project did provide a lift in the capacity of the team leaders in this field. In light of the delay in data entry it might however be reconsidered whether the team leader himself should enter the data next time or whether there could be an alternative. With the exception of Mt. Lebanon, the submission of formal progress reports and updated workplans by the team leaders was not functioning optimally during the fieldwork phase. The information of status and plans for the fieldwork was primarily delivered to the NC-FI verbally. For future repetitions the example set by Mt. Lebanon Mohafaza is recommended followed by all Team Leaders. 15.3 Capacity to validate and merge data The NC-FI was trained by the Technical Backstopping Officer from FORM during September 2004 in the fundamentals of validating and merging data and provided with a crash course in the fundamental statistics behind the NFA. Validating and merging the data from the secondary databases of the team leaders into the primary database at MOA was a time consuming, tedious task which the NC-FI tackled well. Part of the reasons for the time consumption was that some of the RDS’s had used an older version of the database application and therefore the database structure had to be checked before merging. Another cause of time consumption was that although the field forms had been double checked for errors, errors could still occur when entering the data by the team leaders and these needed to be weeded out before processing and analyzing the data. The NC-FI developed his skills in the Access Database application, from user level before TCP/LEB/2903 to highly skilled after the survey. The NFA had therefore elevated the capacity of the NC-FI and thereby MOA. Preferably the skills should be transferred to more individuals within MOA. 15.4 Capacity to process and analyse data The need to recruit a specialist for the highly specialized tasks of processing and analysing the data was recognized at the end of the second project phase.

101

The specialist in data analysis and statistics worked with the NC-ID, which caused a transfer of methods and analytical skills. Since the skills required for data processing and data analysis are highly specialized, it is foreseen that the tasks related to data processing and analysis will need specialist support during future repetitions. A common comment related to all the skills developed during the NFA is that a main challenge ahead is maintaining this level of skills in order to allow for future re-measurement of the permanent sample sites. 15.5 Capacity to produce maps Several factors combined to build the DRDNR capacity for map production substantially during the course of the project. Among the factors of importance for the success of the mapping component the following should be mentioned:

• The presence of a skilled and motivated GIS specialist in DRDNR (Mr. Roukoz) • The positive effect of having a joint national consultancy in remote sensing and mapping

shared by Mr. Roukoz of DRDNR and one of the leading national GIS capacities (Mr. Lichaa) – this caused a transfer of skills to DRDNR and ensured linkage to the Land Use Land Cover Map.

• The willingness of other institutions, especially the Lebanese Army, to provide mapping materials.

During the project the in-house capacity to produce field maps for the field teams was built. The production of the Derived Forest Map, based on the interpretation of satellite imagery, georeferenced maps of forest types, climatic data, ground verification provided by the field work and the subsequent ground validation built the capacity of DRDNR for conducting reinterpretation of images, aided by all relevant existing data. The process of installing and using project equipment as well as the huge work related to producing the working maps and the output maps has implied a substantial capacity building element for the DRDNR NC-RS through learning by doing. Preferably the skills should be transferred to more individuals within MOA. 15.6 Capacity to organize and coordinate an NFA. A review of the responsibilities of the MOA staff involved in the NFA showed that most of the activities required to repeat the assessment are already contained in the job descriptions of the staff of DRDNR and RDS (TAMIM 2004). The NFA therefore provided an opportunity to train in tasks that to a large extent were already contained in the job descriptions of the various staff members. The successful output of the current NFA is proof of a substantial capacity building in all organizational and logistical aspects related to conducting an NFA within the current structure of MOA.

102

16. Time and cost analysis The current section contains an estimate of the consumption of time and transportation for the current NFA. The figures are based on the estimated time consumption on NFA activities during the period July 2003 – September 2005.

Table 54: Estimated MOA working hours spent on NFA Activities (July 2003 – Sept. 2005) Approximate working hours spent on NFA activities

NC Forest Inventory 3,150 NC- Remote Sensing and Mapping 2,300 NC-Institutional Development 2,600 NC-Data analysis 700 Team Leaders 2,700 Forest Guards 5,500 Total 16,950

The efforts of the external staff (IC, FORM backstopping officer, TCDC expert) are not included in the estimates of the work needed for the NFA as they are viewed entirely related to the current project as future repetitions when the NFA will be repeated as a regular task of DRDNR. Considering that a normal working year at MOA lies around 1600 working hours, conducting an NFA is a substantial undertaking especially for the office staff. In excess of 10 man-years have been allocated for the current National Forest and TOF Assessment. Many of the tasks can be considered to run more smoothly for the repetition of the NFA due to the capacity built during the current project and some tasks can be considered to be undertaken only during the current project – e.g. the establishment of the permanent sample sites, the study tours to Guatemala, major workshops, etc. Nonetheless, the next repetition of the NFA will still be a major task that will be requiring the full focus of DRDNR. Concerning the transportation aspects during the fieldworks phase: Five 4WD vehicles were needed for 4 days a week for a total of one year at the RDS’s and the NC FI was allocated access to a 4WD vehicle for the full duration of the NFA. The estimated total number of kilometers driven for the NFA exceeds 36.000 km. For the next NFA it is likely to be less due to more experienced field crews and depending on whether a Derived Forest and TOF map will be produced for the next NFA. Concerning equipment: The equipment mentioned in section 4.4 was purchased for the NFA and will be available for future re-measurements – there will of course be running maintenance and replacement costs. For future repetitions, much of the hardware and software purchased for TCP/LEB/2904 may need upgrading. By next repetition more accurate GPS receivers may be affordable which would be an asset for more accurate location of plot starting points.

103

17. Conclusions TCP/LEB/2903 represents the first national experience with National Forest and TOF Resources Assessment based on systematic field sampling in Lebanon. The NFA is based on the methodology developed by FAO/FORM. The current NFA project should be seen as a pilot study. It provides a first snapshot of the state of the forest resources in Lebanon based on ground measurements. The experience from the current NFA shows that the approach promoted by FAO is a cost efficient way of building national capacity to generate and analyse data on the forest and TOF resources – in many cases the data relates to aspects of Forests and TOF which have not earlier been investigated. The similarity of the area-datasets of the NFA and the DFM at the first and second level of vegetation classification indicate that the NFA area estimates at these levels are representative. By investing about 17.000 man hours in the NFA-related work, a primary set of data on Forests and TOF has been generated, the capacity of DRDNR has been built in all aspects related to forest inventory and mapping and a permanent systematic sampling system has been set up which allows for future monitoring. The data set is broader in scope than earlier studies of the Lebanese Forest and TOF resource and it is based on a systematic sampling grid and a methodology that allows for future repetitions and expansions of scope. With repetitions, the NFA will capture information related to the change in the extent, state and uses of the forest and TOF resources and it will be possible to monitor the changes. Additional parameters can be included for future measurements if deemed relevant. The value will therefore increase with every re-measurement. The permanent sampling grid could offer a platform for collaboration with other institutions for expanding the scope and increasing the level of detail of the NFA. The NFA should be seen as a starting point for a tradition of conducting systematic inventories and using the gathered information on the forest and TOF resources in the development of appropriate national policies and plans and their subsequent implementation, i.e. a continual process as illustrated by the below figure.

Figure 19: NFA and policy development

The ultimate value of the NFA will depend on how useful the data captured is. It is therefore important to follow up on the NFA constructively in order to ensure that the data captured is relevant for meeting national needs. Developing the correct policies depends on timely and relevant

104

information. The NFA is an important step for providing such information and will continue to evolve as the knowledge of the Forest and Tree Resources continue to evolve. The NFA should therefore not be seen as the end product in itself – but the start of a process. The results of the NFA show significantly higher figures for the areas covered by Forest and TOF than previously reported. These findings can be instrumental in a debate about the extent of knowledge about the Forest and TOF Resources and the importance of access to timely and relevant data on the Forest and TOF resource as well as a catalyst for more research in this field. Conducting the NFA has been a two-way learning process, for the staff and institutions involved and also for the FAO, as NFA Lebanon has generated valuable feedback to FORM on the promoted methodology.

105

18. Recommendations 18.1 Future Monitoring of Resources. MOA/DRDNR decided on a repeating the NFA as a periodic inventory with the aim of repeating the NFA every 5 years. It is therefore intended to undertake the work in a single year which is expected to imply a more efficient organization, supervision and control of the work as compared to conducting it continually. It is also expected to be easier to allocate MOA resources to perform periodic NFA’s and it will make the importance of the work will more visible. MOA recommends that the field teams in the future should still be composed of forest guards as they are field based, they have the local knowledge and they are part of the structure of DRDNR. The field teams should be headed by a team leader, normally the Regional Chief of Service RDNR.

It is recommended that field work should be at the level of 2 working days /week in the years that the fieldwork is conducted. Field manual and field forms should be translated to Arabic in order to involve a broader group of qualified forest guards. This would permit the Field Team Leaders to perform their regular duties as well at the Regional Development Services. The nature of the future inventories influences the planning of the work of DRDNR for the coming years with respects to institutional development needed to embed the NFA in the regular tasks of the DRDNR as well as for the collaboration with other institutions. Below is a proposal presented by the NC-ID for a scheduled repetition of the NFA.

Table 55: Proposal for NFA repetition

2006 On the job training and maintenance of the skills for field crew and all staff involved in the NFA

2007 On the job training and maintenance of the skills for field crew and all staff involved in the NFA

Request for needed mapping material and expertise from relevant institutions Coordination between Regional Development services and DRDNR Coordination with other institutions if they want to assist in the NFA (July – December 2007)

2008 Construction of Fieldwork Maps (December 2007- March 2008).

Preparation of field measurement equipment (January – February 2008)

Logistical preparation for field crews (January – March 2008)

Identifying volunteers e.g. from NGOs or Universities (January- March 2008) Establishing field crews Data Collection and Data Entry (April 2008 – June 2009)

2009 Data Collection and Data Entry (April 2008 – June 2009)

Data Processing and Analysis, Reporting. (June - Dec 2009).

2010 Reporting 2010.

(Source: TAMIM 2004)

106

Concerning the allocation of responsibilities for future National Forest and TOF Assessments, the NC-ID proposed the following:

Table 56: Proposed allocation of responsibilities for future NFAs

Task Responsible Overall Supervision and responsibility Director DRDNR

Supervision and Control of Data Collection and Data Entry

Head of Service Reforestation & Exploitation (DRDNR/MoA)

Data Processing and Analysis Specialist in Data Analysis and Data Processing (DSC/MOA or external) in collaboration with Head of Service Reforestation & Exploitation (DRDNR/ MoA)

Mapping and Remote Sensing GIS Unit, DSC/MOA

Reporting DRDNR FRA Correspondent

Field work 1Field crew: 1 team leader (Chef Department DRDNR in the region) and 3 forest guards:

(Source: TAMIM 2004)

A review of the responsibilities of the MOA staff involved in the NFA shows that most of the activities required to repeat the assessment are already contained in the job descriptions of the staff of DRDNR and RDS (TAMIM 2004).

To provide overall guidance and to facilitate inter-institutional collaboration, the NC-ID recommended that a Steering Committee or Advisory Board could be incorporated into the future structure of the NFA The NC-ID recommended that DRDNR, Service of Reforestation and Exploitation should include the NFA in the plan for the year 2008-2010 along with the budget needed in order to cover:

• provision of services by other institutions and material to be purchased. • extra working hours for the field crews (time after 2:00 p.m.). • fuel and maintenance for vehicles assigned for the NFA.

In order to maintain the skills over time for staff at all levels, the NFA training should be ample and relevant. The training should therefore aim at finding and re-measuring the NFA tracts in accordance with the field manual– this will provide realistic training on re-measurements including locating the starting points with the metal locator. Part of the training could also aim at measuring a tract in vegetation types were the data collected through the NFA is sparse or missing – e.g. in fir forest. The data collected through the training sessions will not be a part of the NFA database, but can supplement the NFA data in areas where data is missing e.g. standing volume of fir, for reporting purposes.

107

Allocation of sufficient time and resources for securing that all data in the database is double checked and entered correctly is important for the future NFA as the quality of the NFA output depends on the quality of the NFA input. Data validation / double-checking should be allocated sufficient time and resources. For future NFA’s an additional tract in the extreme south of Lebanon near Rmaich village (not included in the original sampling grid provided by FORM) should be included. 18.2 Additional Information Needs. The NFA provides new knowledge on the forest and TOF resource of Lebanon. It also helps identify areas that should be studied further. Among areas that could be investigated further are the following: � Development of more appropriate methods of measuring growing stock in orchard trees. � Field measurement of vegetation classes that were not captured by the field inventory (e.g. fir). � Forest and TOF maps (e.g. detailed mapping based on aerial photos and maps for management

purposes). � Studies for elaboration of volume in dense shrubs and coppices � Additional studies of Products and Services from Forest and TOF for management purposes. � Additional species and location specific studies in the relationship between age and growing

stock It applies that many of the areas for potential further study could be subject to institutional collaboration to gather data of sufficient level of detail for management purposes to ensure sustainable utilization. 18.3 Actions for Resources Development The rural presence of the forest centres and the fact that the DRDNR through forest guards is present throughout the country makes DRDNR well positioned to undertake the future systematic monitoring of the resources. The high level of skills among the forest guards, technicians and engineers, their local knowledge as well as the recent NFA data are a good foundation for developing the resources. The value of the NFA data will increase with every re-measurement of the permanent sample sites in order to capture change information, and the data and the permanent sampling grid is potentially useful to many other institutions. Collaboration concerning the fieldwork and the sharing of data between relevant institutional stakeholders should therefore be promoted. It could also help spreading the workload related to conducting periodic fieldwork. The project findings will help identify priority areas and be of use for the development of a forestry action plan and an analysis of forestry policy/strategies in order to ensure a coherent legislation. With reference to the illustration in section 17 the NFA is part of a continual process and not the end product in itself. The large share of private property in Lebanon and the high density of population makes it recommendable for the development of successful policies or management plans for Forests and

108

TOF to include a participatory approach with the private stakeholders (BEYDOUN AND STEPHAN 2005). The development of land zoning and management plans e.g. for communal land was identified in the January 2005 project workshop as a possible area of focus along with the importance of linking forest management to rangeland management policies and the need for distinction between legislation related to forest areas and to protected areas. 18.4 International Reporting The NFA will facilitate International Reporting as it produces the figures required by a number of international processes. The driving force behind the NFA is however the extent to which data is of use at national level e.g. for policy development or for resources management. A consequence of conducting sample-based inventory is that some vegetation types of rare occurrence may not be captured by the inventory. In the case of the NFA fir was not captured. The area figures to be used for reporting purposes should therefore always be taken from the mapping component of the project, as this is the most accurate. The data related to the composition, state and uses of the Forest and TOF Resources will always come from the field inventory as the only way to capture reliable information is through on the ground measurements. The National Forest and TOF Resources Assessment of Lebanon is a two-way process, building national capacity in NFA and providing continuous feedback on methodology to FAO / FORM. Of concise recommendations from the current project are the following: • It is recommendable to improve the compatibility between the different classification systems

employed by FAO, i.e. the land use classification systems of FRA and LCCS.

• It could be considered to include age measurements in the base parameters, to include laser rangefinders in the standard equipment list and to supplement the measurements with a more objective way of estimating density and species distribution (e.g. basal area counts).

• It could also be considered to make the soil parameters in the field forms optional - in the case where countries already have good soil maps, as is the case in Lebanon.

109

19. Bibliography ABI SALEH, B., SAFI, S., 1999 Map of Vegetation levels. Republic of Lebanon, Ministry of Environment. ASMAR, F. R. 2003. National Report on the Forestry Sector in Lebanon. April 2003. Ministry of Agriculture, Republic of Lebanon. Unpubl. BASSIL, M. 2003. Presentation for TCP/LEB/2903 Steering Committee, December 2003. Unpubl. BASSIL, M. 2004. Progress Report – National Consultant Forest Inventory TCP/LEB/2004. Unpubl. BEYDOUN, G. Y. AND STEPHAN J., 2005. National Forest Assessment Program Lebanon. Data Analysis Report by Genane Youness Beydoun – FAO Data Analysis Expert & Jean Stephan Agricultural Engineer, MOA. CONSOLACION, C.P., 2004. TCDC Consultant’s Final Report. National Forest and Tree Inventory and Assessment. TCP/LEB/2903. Ministry of Agriculture, Lebanon / Food and Agricultural Organization of the United Nations. Beirut, July 2004. Unpubl. DALSGAARD, S., 2003A. International Consultants Report no. 1. TCP/LEB/2903 National Forest and Tree Inventory and Assessment, Lebanon. Ministry of Agriculture, Lebanon / Food and Agricultural Organization of the United Nations. Beirut, August 2003. Unpubl. DALSGAARD, S., 2004A. International Consultants Report no. 3. TCP/LEB/2903 National Forest and Tree Inventory and Assessment, Lebanon. Ministry of Agriculture, Lebanon / Food and Agricultural Organization of the United Nations. Beirut, April 2004. Unpubl. DIRECTORATE OF GEOGRAPHIC AFFAIRS, LEBANESE ARMY, 1965. Forest Type Map. Issued 1965 by Forestry Education, Training and Research Project. Republic of Lebanon – Green Plan (K. El Husseini) United Nations Special Fund (FAO). FAO 2000. Land Cover Classification System. LCCS. Classification Concepts and User Manual. A. Di Gregorio, FAO Environment and Natural Resources Service, L.J.M. Jansen FAO Land and Water Development Division. Rome 2000. FAO 2001. Global Forest Resources Assessment 2000. FAO Forestry Paper 140. Rome 2001 FAO 2004A. FAO Forest department. FRA website, country brief for Lebanon FAO 2004B. Global Forest Resources Assessment update 2005. Guidelines for Country Reporting to FRA 2005. Forest Resources Assessment Programme. Working Paper 82. Rome 2004 FAO 2004C Projet de Recensement Agricole & Ministry of Agriculture/Directorate of Studies and Coordination, 2004. Agriculture in Lebanon 2003. LICHAA, EL-KHOURY, D. & BAKHOS, W., 2002. Land Cover – Land Use Development Program Draft Technical Report. Lebanese Environmental and Development Observatory (LEDO) Ministry of environment (MoE), Ministry of Agriculture, MoA, Lebanese National Council for Scientific

110

Research (LNCSR), National Council for Remote Sensin (NCRS) Council of Development and Reconstruction (CDR)represented by the IAURIF (Institut d’Aménagement et d’Urbanisme de la Région Ile-de-France). Beirut May 2002. Unpubl. MINISTERE DE LA DEFENSE NATIONALE, 1965. Carte du Liban, 1:20.000. Etat-major de l’armee. Direction des Affaires Géographiques et Géodésiques. Beyrouth, Liban. MOA 2003. National Action Program to Combat Desertification, Ministry of Agriculture, Lebanon. June 2003. MOA/FAO 2003. TCP/LEB/2903 Project Document. Ministry of Agriculture, Lebanon / Food and Agricultural Organization of the United Nations. Unpubl. Mouterde P.S.J., 1966-1980. Nouvelle flore du Liban et de la Syrie. 3 tomes textes, 3 tomes atlas. Dar el Machreq, Beyrouth. NFA 2005, Guidelines for technical reports ROUKOZ, S. 2005. Final Report - National Consultant Mapping and Remote Sensing TCP/LEB/2903. MOA, Beirut, July 2005. STEPHAN, J. 2005. Presentation of fieldwork experiences. TCP/LEB/2903 Technical Workshop, Beirut . January 2005. Tamim, Z., 2004. Final report – National Consultant Institutional Development, TCP/LEB/2903, MOA, Beirut December 2004.

UNEP & Ministry of Agriculture, 1996. Biological diversity of Lebanon (9 volumes).

111

Appendix 1: Statistical functions used in data analysis The following functions were used (BEYDOUN & STEPHAN 2005):

Ratio estimate

® ��

�

�=== n

i

ni

n

i

ni

y

x

yn

xn

yx

1

1

1

1

1

1

Standard deviation of population

(Sd) ( ) ( )

( ) ( )1

2*1

1

11

2

2

1

2

11

2

−

−+

==

���

��

��

n

y

x

y

x

V

n

iin

i

n

in

in

i

n

in

i yxyx

Standard error of Ratio estimate

(SR) ( ) ( )

( ) ( )( ) ( )1*

2*

* 2

1

1

11

2

2

1

2

11

2

−

−+

==�

���

��

��

ny

y

x

y

x

nVi

n

iin

i

n

in

in

i

n

in

i

R

yxyx

Sampling error

(SER) ( ) ( )

( ) ( )( ) ( )1*

2*

** 2

1

1

11

2

2

1

2

11

2

−

−+

==�

���

��

��

ny

y

x

y

x

ntSti

n

iin

i

n

in

in

i

n

in

i

R

yxyx

By assuming a normal distribution for the ratio a normal deviation t

2α =1.96 with a significance

level �=5% was used Relative Sampling error

(SER%)( ) ( )

( ) ( )

( ) ( )1**

2**

1001002

1

1

1

11

2

2

1

2

11

2

1−

−+

==��

���

��

��

�nyx

y

x

y

xn

ytR

SE

i

n

i

n

iin

i

n

in

in

i

n

in

i

n

iR

yxyx

For the strata the following apply: Total sample of population, total sample size

(nT) 22621 ... nnn +++= And nk=2 ha , for every k=1,….,226

112

Total population size (NT) 22621 ... NNN +++= And Nk= 4625 ha, for every k=1,….,226 So that the weight for each stratum ratio estimate is 1/226 Estimate of overall Ratio estimate

(RT)( )

���

���

�

=

== =��

��

�

==226

11

11

1

1

1

**

jn

ji

nji

k

jn

ji

nji

j

k

jjj

y

x

N

y

xN

N

RN=R

Estimated Total

(XT) ( ) ���

�==

��

��

�

===

k

jn

ji

n

jij

k

jjjTT j

j

y

xNRNRN

11

1

1

*** =Nt*R