1 REPORT OF THE MID-TERM WORKSHOP COMPONENT 1 FFEM FFEM PROJECT "MAXIMIZE THE PRODUCTION OF GOODS AND SERVICES OF MEDITERRANEAN FOREST ECOSYSTEMS IN THE CONTEXT OF GLOBAL CHANGES" COMPONENT 1 "PRODUCTION OF DATA AND DEVELOPMENT OF TOOLS TO SUPPORT DECISION AND MANAGEMENT OF VULNERABLE MEDITERRANEAN FOREST ECOSYSTEMS AFFECTED BY CLIMATE CHANGE AND THE ABILITY OF THESE FOREST ECOSYSTEMS TO ADAPT TO GLOBAL CHANGE" Tunis, Tunisia 2-5 June 2014
59
Embed
REPORT OF THE MID-TERM WORKSHOP COMPONENT 1 FFEM … · LoA: Letter of Agreement (MOU) MENA: ... DGF Tunisia Youssef Saadani [email protected] GIZ Tunisia Abdelmajid Jemaï [email protected]
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
REPORT OF THE MID-TERM WORKSHOP
COMPONENT 1 FFEM
FFEM PROJECT "MAXIMIZE THE PRODUCTION OF GOODS AND SERVICES OF
MEDITERRANEAN FOREST ECOSYSTEMS
IN THE CONTEXT OF GLOBAL CHANGES"
COMPONENT 1 "PRODUCTION OF DATA AND DEVELOPMENT OF TOOLS TO
SUPPORT DECISION AND MANAGEMENT OF VULNERABLE MEDITERRANEAN
FOREST ECOSYSTEMS AFFECTED BY CLIMATE CHANGE AND THE ABILITY OF
THESE FOREST ECOSYSTEMS TO ADAPT TO GLOBAL CHANGE"
Tunis, Tunisia
2-5 June 2014
2
This report was drafted by Sophie Valley (AIFM) under the LoA of 26.04.2013 signed between AIFM
and FAO. It was then validated by team of FAO (V. Garavaglia and C. Besacier).
health problem: bark beetle attack (cleaning process from 1989 to 1992)
Steps implemented since the workshop of Solsona (May 2013) to define a methodology for a vulnerability analysis of Mediterranean forest ecosystems in the pilot sites
Selection of national experts Gacemi Mohamed El-Amine (Centre of Space Techniques - CTS)
Signatures of MoUs/ agreements with FAO
No contract signed with the local structure. Expenses incurred directly by the Permanent Representative of FAO in Algiers (Field Budget Authorization)
Others Coordination meeting at the DGF for the launch of the project (C1 and C4). Support mission on the pilot site of Senalba along with experts (FAO and ONFI). Review and discussion of the methodological guide for component 1 (Part I) with the FAO expert on data availability and existing inventories. Adoption of the terms of reference related to health aspects of forests and recruitment of a consultant to support the management of this component within the CTFC
First data collection for analysis of vulnerability to climate change:
Physiographic and forest data
Map of site limit and forest/non-forest limit Paper charts of the forest Inventory of 1984 (to be digitized) Maps of land cover (1987, 2001, 2011) Map of forest types Map of aging Map of soil depth, soil texture map Geological map Maps of altitudes, slopes, drainage networks, exhibitions Map of Fires Map of desertification sensitivity (CTS, 2010) Forest Inventory 2008
Meteorological data Rainfall, temperature (AGRI4CAST) T min, T max, HR, P, etc. (Tutiempo.net)
Demographic and socio-economic data
Population density, livestock
Satellite images Google, Microsoft Bing (good resolution)
First results, if available Change of the Pluviometric Quotient of Emberger since 1975 and overlaid on Emberger's Statistical Graphs -> Since 1975 the region of Djelfa experienced 3 bioclimatic stages
Training in collaboration with GIZ Tunisia on "Multifactorial space approach for the analysis of vulnerability of ecosystems of Component 1 pilot sites to face climate change" (21-25 April 2014):
Applicability in the pilot site Yes technically
Problems Parameters, coefficients and weights used in this approach change from one area to another, and require the opinion of several experts every time
The factors used are not the same (no fire, but strong trend towards desertification)
Possible Solutions Tool for calculating vulnerability dynamically and automatically
Training in collaboration with VITO about "Cartographic material on the evolution of forest cover, land use and climatic conditions of the pilot sites" (19-23 May 2014)
Reference data available Forest inventory maps 1984
Problems Use of the software Spirit
Possible Solutions Additional training on the software Spirit or tutorial
Next steps 1 - Formatting data collected and their standardization following the multifactor spatial analysis approach of the ecosystems vulnerability of GIZ 2 - A field mission for:
studying some vulnerability factors with forestry experts on the pilot site
updated maps of forest inventory 3 - Application of the multifactor spatial analysis approach with the available data 4 - Analysis of Landsat data downloaded 5 - Formatting downloaded weather data 6- Production of documents on the evolution of forest cover, land use and climate
Budgetary constraints/ budget situation
The budget was approved by both parties. No constraints at this stage for the implementation via FAO Algiers
11
Pilote site of Jabal Moussa (Lebanon) Miguel Ángel Navarrete Poyatos
Notes on the site
Location
Qadaa of Keserwan, 45 km north-east of Beirut western side of Mount
Lebanon
Altitude 750 -1 250 m
Surface 6500 ha including 1250 ha of woodland
Population on the site About 15 000 inhabitants, with 1 500 users of woodlands
Notes on the site Biosphere Reserve (UNESCO)
Bioclimatic stage Sub-humid and humid for lower areas (Thermo-and Eu-
Mediterranean)
Humid temperate in medium altitude areas (Supra-Mediterranean)
Per-humid fresh at the highest peaks (Montagnard or
Mediterranean-mountain)
Geology/Soil science Calcareous soils
Ecosystems origin Forest series, from low to high altitude: Quercus calliprinos,
Species present Presence of rare and / or endemic species: Asperula libanotica, Cyclamen
libanoticum, Origanum libanoticum, Pentapera sicula var. libanotica, Malus
tribolata, Acer tauricolum
Evaluation of goods and
services:
- Benefits for direct use
(socio-economic activities)
Wood products (timber, firewood)
Charcoal
Agroforestry (fruits)
Aromatic and medicinal plants
Silvopastoralism
tourist activities (ecotourism, hiking)
Nursery seedling producing local
12
- Benefits for indirect use Recreational space
Landscape, cultural and historical
Pressures from human
activities
Extraction of material (rocks, sand)
Construction/urban/infrastructure
Overgrazing (related to the number of animal data)
Coal (currently regulated by the Ministry of Agriculture through
permits coal)
Significant tourist traffic in the summer
Fires
Owner (s) of woodlands Macha'a (commons)
Religious WAAF
Family WAAF (Zouaine)
Private owners
Bank of Lebanon
Manager (s) of woodlands Municipalities
Religious institutions
Private owners
Association for the Protection of Jabal Moussa (APJM)
Health of forests Mistletoe on juniper trees
Steps implemented since the workshop of Solsona (May 2013) to define a methodology for a vulnerability analysis of Mediterranean forest ecosystems in the pilot sites
Selection of national experts
2 experts recruited: Miguel Ángel Navarrete Poyatos as a national expert
and Salim Roukoz as a support for GIS
Signatures of MoUs/ agreements with FAO
LoA signed between FAO and APJM (Association for the Protection of
Jabal Moussa). Contracts signed between the Ministry of Agriculture of
Lebanon and APJM
Others Regular meetings between APJM, MoA and the experts
First data collection for analysis of vulnerability to climate change:
Physiographic and forest data
Forest Inventory (random sampling, stratified - 2012) Map of models of forest fuels Land use maps (LULC - 2002, 2005) Geological map, map of soil types, soil depth map Maps of altitudes, slopes, drainage networks, exhibitions Map for risk of erosion and landslides
Meteorological data 2011-2014 monthly data series: solar radiation, wind direction (degrees), precipitation, wind speed, leaf wetness, air temperature, relative humidity, dew point, soil temperature, air pressure, reference evapotranspiration
First results, if available Delimitation of population, crossing layers of slope, aspect, type and soil depth and altitudinal range (threshold of 1000 m)
Evapotranspiration, available water capacity
State of health of populations obtained through remote sensing techniques
Potentiality species to current and future climate conditions
Map of fire risk potential by combining forest slope and fuel
13
type
Training in collaboration with GIZ Tunisia on "Multifactorial space approach for the analysis of vulnerability of ecosystems of Component 1 pilot sites to face climate change" (21-25 April 2014):
Applicability in the pilot site Yes for some factors, but not for others
Problems Problems for some risk factors
Possible Solutions Biophysical factors: No change
Water stress: No data on vegetation stress => Change to the
potential of species to climatic factors
Ageing: No data on the age of trees => Change to the health of
forests
Grazing pressure: No data of pastoral inventory, impossible to
estimate forage units => Ignore this factor (factor interrelation
with forest fires)
Risk of forest fires: No change
Training in collaboration with VITO about "Cartographic material on the evolution of forest cover,
land use and climatic conditions of the pilot sites" (19-23 May 2014)
Applicability in the pilot site Partially.
Quotient Emberger and water stress indicator (ASIS) will be calculated.
Impossible to work with the data MARS available
Landsat images downloaded for free
10-15 images of 1990, 2000 and 2010.
Work for GIS and estimation of the Normalized Difference Vegetation
Index (NDVI) in progress
Reference data available Points of presence by species (vector files) to form the model.
Classification of vegetation in progress
Problems Impossible to work with the set of proposed climate data (March) due to
its spatial resolution (25x25 km). A pixel is larger than the entire
Biosphere Reserve of Jabal Moussa.
Possible Solutions Modelling with BIOMOD2 (see note below and Appendix 2)
Next steps 1. Development of a synthetic factor of vulnerability according to
multivariate spatial modelling
2. Classification of vegetation by Remote Sensing
3. Calculate quotient Emberger and water stress indicator (ASIS).
Analysis of the evolution of the NDVI
4. Preparation of final results
Budgetary constraints/ budget situation
First transfer received. Waiting for the second transfer. Budget lines met.
Uncertainties regarding the cost of data to be collected, it will be
quantified in the following months.
14
Modelling with BIOMOD2 (Thuiller et al., 2013), an additional package used in the statistical software
R, allows the modelling of current and future potential distribution of species. The modelling requires
two types of data: climatic variables (past and current) and points of presence/absence of the
species. The idea is to identify the areas with favourable conditions for the presence of a species.
In the case of Lebanon, climate data were issued from WorldClim (resolution 1 km²). The selected
scenarios of climate change are those of the IPCC: A2 (pessimistic) and B1 (optimistic). The horizon
predicted is 2050. Collinearity analysis was performed in order to eliminate redundant variables (with
a strong correlation). Three non-collinear variables were selected: winter minimum temperatures,
summer maximum temperatures and summer precipitation.
15
Pilot site of Maa mora (Morocco) Fouad Mounir and Belghazi Bakhiyi
Note on the site:
Location East of Rabat and Kenitra, the western part is 10 km from the ocean
Altitude 7 m West and 290 m to the East
Surface 132,000 ha of which 126,200 ha of woodland
Population on the site approximately 341 360 hab. including 173,500 users of woodlands
Notes on the site largest cork oak lowland in the world
Bioclimatic stage subhumid in the Western part
semi-arid in Central and Eastern
Geology/Soil science sandy soils called "beige" soil (Quaternary dunes on Miocene marl)
Ecosystems origin cork oak (Quercus suber) in the Maamora Poirier (Pirus mamorensis)
Ecosystems currently present
Quercus suber in the Maamora Poirier (Pirus mamorensis)
plantations of non-native species (eucalyptus, various
pines, acacia)
Species In addition to those mentioned before: Pistacia lentiscus, Pistacia
Steps implemented since the workshop Solsona (May 2013) to define a methodology for vulnerability analysis of Mediterranean forest ecosystems in the pilot sites
Selection of national experts 2 experts recruited: Bakhiyi Belghazi and Fouad Mounir
Signatures of MoUs/ agreements with FAO
Agreement signed with a local structure
First data collection for analysis of vulnerability to climate change:
Physiographic and forest data
General forest data, data on forest fires
History of forest management, forest crimes
General physiographic data with clay floor, deep sand, slope map
Demographic and socio-
economic data
Total population, households, population users, livestock
Identification of socio-economic sectors
Data on cork harvesting
First results, if available Assessment of available soil resources, water deficit
Analysis of the dynamics of Maamora forest (1990-2010)
Assessment of acorns seedling and plantations of cork oak
Training in collaboration with GIZ Tunisia on "Multifactorial space approach for the analysis of
vulnerability of ecosystems of Component 1 pilot sites to face climate change" (21-25 April 2014):
Applicability in the pilot site Yes
Training in collaboration with VITO about "Cartographic material on the evolution of forest cover,
land use and climatic conditions of the pilot sites" (19-23 May 2014)
Applicability in the pilot site Yes
17
Pilot site of Siliana (Tunisia) Kamel Tounsi and Ali Aloui
Note on the site:
Location Siliana region, about 120 km southwest of Tunis
Surface 91 000 ha of which 23 500 ha forest areas
Population on the site Around 11 600 hab., 8 300 forest users
Bioclimatic stage Semiarid
Ecosystems origin Aleppo pine forest
Ecosystems currently present
Summit: Montpellier maple
Mid-slope: low and dense scrub of Aleppo
Pine and green oak
Low slope: garrigue with oleaster, mastic, rosemary,
Cistus
Evaluation of goods and services: Economic evaluation of goods and services of Tunisian forests
(Daly et al., 2012)
- Benefits for direct use
(socio-economic activities)
Wood products (wood work, wood industry , firewood)
Aromatic and medicinal plants , extraction of essential
oils
Collection of mushrooms
Various crops: snails, acorns, asparagus, Aleppo pine
cones, pods, etc.
Beekeeping
Forage production
Hunting activities
- Benefits for indirect use Recreational space
Biodiversity reserve
Reduction of sedimentation in the dam
Carbon sequestration
Protection against erosion
Pressures from human activities Grazing (pastoral overload)
Anthropic pressure (many crops, woody and non-
woody)
Owner (s) of woodlands State and local communities (communal lands)
Manager (s) of woodlands State (DGF) and user population
Current forest management Conservation, protection and regeneration of mountain
maples
Preserving, protecting and management of oaks for
the production of acorns and wood
Introduction of animal species that used to populate
18
the mountains, like the Barbary sheep and Cuvier's
gazelle
Environmental education for school
Structure and organization of the population through a
participatory approach, in view of improving their
quality of life through the development and the
development and implementation of development
plans
Steps implemented since the workshop Solsona (May 2013) to define a methodology for a vulnerability analysis of Mediterranean forest ecosystems in the pilot sites
Selection of national experts 2 experts recruited: Ali Aloui and Kamel Tounsi
Signatures of MoUs/ agreements with FAO
MoU signed between FAO and Development Association
Menzel Jemil (ADMJ)
First data collection for analysis of vulnerability to climate change:
Physiographic and forest data Pastoral and forest inventories 1995 and 2005
Thematic maps: soil science, slopes, exhibitions
Map of the sectors (administrative division)
Map of bioclimatic zones
Thematic map of altitudes classified by classes 100 m
Fire Statistics 1985-2014
Phyto-ecological map of Northern Tunisia – scanned
(C. Floret, J.L. Guillerm, E. Le Floch and A. SRouler–
1966).
Meteorological data Climatic data: rainfall, temperatures for Siliana, Kef, Jendouba,
Beja, Sidi Bouzid,
Demographic and socio-economic
data
Data from 2004 and earlier populations + projection
2034
Agricultural Statistics 1995 and 2005 (livestock) and
projection by OEP
Satellite images Google Earth
First results, if available 1. result: the choice of spatial reference unit: the
Juniperus oxycedrus L. Prickly juniper/Genévrier oxycèdre
Juniperus phoenicea L. Phoenician juniper/Genévrier rouge
Laurus nobilis L. Bay laurel/Laurier noble
Pinus brutia Ten. Turkish pine/Pin de Calabre
Pinus halepensis Mill. Aleppo pine/Pin d'Alep
25
Pinus nigra J.F. Arnold European black pine/Pin noir
Pinus pinea L. Stone pine/Pin pignon
Pistacia lentiscus Mastic tree/Lentisque
Platanus orientalis L. Oriental plane/Platane d'Orient
Quercus coccifera L. Kermes oak/Chêne kermès
Quercus cerris L. Turkey oak/Chêne de Turquie
Quercus ilex L. Holm oak/Chêne vert
Quercus canariensis Willd. Algerian oak/Chêne Zéen ou zen
Quercus suber L. Cork oak/Chêne liège
Taxus baccata L. European yew/If
Tetraclinis articulata (Vahl) Mast. Barbary thuja/Thuya de Berberie
The purpose is to produce maps of the 24 native species distribution and determine the marginal
populations. The approach is the following:
1. Collection and compilation of all available data
2. Screening and optimization of selected data
3. Production of maps
GIS software used: Q-GIS
Collection of data: 3 data types:
points of presence (entered by users or different observers),
geographic coordinates (X, Y) of actual populations in scientific publications,
online shapefiles on the distribution of species (including NFI maps).
Collection: Data sources:
GBIF online database (Global Biodiversity Information Facility): information on the presence
of only one species (and not on the density), based on observations not always scientific of
users. This may be also seeds, fossils, herbarium ...
META database (INRA-Avignon, compiled by Cyrille Conord): info on populations until 2009
EUFORGEN database (European Forest Genetic Resources Programme): maps of distribution
of only a few species, including those from Europe.
How to filter information on autochthone distribution (native distribution) of selected species?
Several sources, more or less reliable, were consulted: Catalogue of Life, Med Checklist, KEW World
Checklist and EURO+MED. Many problems were encountered: the information is not consistent or
specific, often incomplete and sometimes contradictory.
Some maps are available: Quercus suber, Pinus halepensis, Cedrus libani, Juniperus phoenica.
The distribution areas defined by EUFORGEN are then compared with the actual data (data
digitization national papers).
In the case of Quercus suber for example, problems of resolution and therefore the detail of the
information arise: are there wooded areas of Quercus suber that connect the various cork oak
26
landscapes between them? What is the interest to keep such detailed information? From when
(density, area) the information should be available?
The pilot site of Siliana (Tunisia) illustrates this example: there are approximately 1 000 individual
Quercus pubescent. This small isolated patch is bonded to a microclimate, this is a marginal
population. This geographical marginality has a genetic interest in Tunisian forest.
It is important to have the two types of information: large areas where populations are dense and
geographically marginal populations
The next steps are:
Feedback from national experts (important to determine the exact regions with the
presence of the species in a country, also useful for the distribution of native species);
Consultation of national forest inventories (e.g. forest inventory 2008 in Algeria)
Scanning paper maps remaining before September 2014;
Updating the database META by June 2014 (2009-2014 - Susana Pereira - INRA
Avignon);
Latest comments of national experts;
Creation of maps;
Provision of maps (shapefiles and jpeg) to the scientific community by November
2014.
This work will localize marginal and peripheral populations and identify populations at risk, hence
the importance of having detailed and accurate maps.
In addition, this work allows to link with Component 2: What is the impact of changing woodland
(distribution areas of different species) on the provision of goods and services?
II) Next steps of the study conducted by INRA Avignon - Challenges and actions to be
implemented/data to be mobilized to achieve the objectives of this study by the end of
2014
Country Referent Information to be transmitted Date
Algeria Assia Azzi
National Forest Inventory
Pinus halepensis reforestation green dam (period: seventies)
Morocco
Fouad Mounir NAFLO database/some forestry inventory maps of 1994
from 9 to 13 June 2014
Bakhiyi Belghazi Map of vegetation After June 15, 2014
Fayçal Benchekroun National Forest Inventory (map /shapefile)
27
Remarks:
As a result:
There is an urgent need to collect shapefiles data with accurate information on the
source and date of the data.
The results of this study will be presented on maps at Mediterranean-basin level and
not at national level.
It is necessary to have consistency for the 24 selected species, with the same level of
detail (e.g. case of cork oak).
When there is enough information for a species, it will be possible to produce more
detailed maps.
It is necessary to distinguish the area of distribution of the species and the actual
presence of the species.
With respect to maps of natural distribution of species, INRA expects the accuracy of
national experts. For example, the Pinus Pinea was planted in many places in Tunisia,
and the Pinus Halepensis has often been planted in the Mediterranean in areas of cork
oak.
Different data types are employed: bibliographic data, paper maps, forest inventories.
Can we illustrate all these data if we mix them? Can we mix data with different dates?
Algeria: The last forest inventory in 2008 could give a lot of new information. However, it is rather an
administrative forest inventory made by the province and not by forest groups.
Tunisia: Many data are available at the CEFE-CNRS of Montpellier (including data on species in
shapefiles and "map phytoecological of Northern Tunisia" in 1967). However, some data (including
the study on cork oak and degradation in Tunisia) are available in paper charts and need to be
scanned.
Morocco: Maps of species distribution by Emberger in 1939 should be scanned ("Overview of
vegetation in Morocco"). This study is quite complete for common forest species, but there is little
information available for most secondary species such as Arbutus Unedo or Chamerops Humilis.
Lebanon: Data should not exceed the framework of the project.
Tunisia Ali Aloui National Forest Inventory (map/shapefile)
Phytoecological map of Northern Tunisia in 1967
Lebanon
Miguel Angel Navarrete Poyatos
Absence/presence of species in Lebanon
Salim Roukoz Shapefile forest 1965 from 5 to 13
June 2014
Shapefile land cover 2005
Turkey Murat Turkes
Forest Atlas of Turkey (website) from 9 to 13 June 2014
Data and shapefiles for the pilot site
Shapefile data of forest atlas
28
SESSION 3: STATE OF THE ART ON STUDIES/RESEARCH FOR THE ADAPTATION AND MITIGATION OF CLIMATE CHANGE IN MEDITERRANEAN FOREST ECOSYSTEMS: FIRST RESULTS AND WORKPLAN 2014 Valentina Garavaglia
The bibliographic collection made by FAO (Carolina Gallo Granizo) of all studies/researches on
adaptation and mitigation to climate change in Mediterranean forest ecosystems will allow to
establish a basis for inter-component data, which will be available on the webpage of the FFEM
project /Silva Mediterranea (http://www.fao.org/forestry/82782/en/), with a link to the platform
FFEM of Plan Bleu (http://planbleu.org/en/activites/foret/optimiser-la-production-de-biens-et-
services-par-les-ecosystemes-boises).
For each publication there will be a summary table containing: Title, Author, Publication Date, Name
of the journal (if applicable), Volume (for review), Publisher, Language, Study area, Species studied (if
applicable), Component(s) concerned, Pilot site(s) concerned, Keywords, Abstract.
Example of the summary table for each publication
The database online will include search engine keywords (component, country, species).
Regional workshop to support the implementation of methodological approaches
aiming at discussing difficulties encountered and methodological issues still to be
clarified.
How to assess the social and economic value of ecosystem services? Thanks to the cost-benefit
analysis (Cost-Benefit Analysis - CBA). The Cost Benefit Analysis (CBA) is a method for the
description and aggregation of the expected effects of a decision. It is an analysis with respect to an
investment and not with respect to a good. CBA will help deepen the existing analyses and thus go
further in the production of useful and relevant information to the decision:
What are the changes in the economic and social value of goods and services based on
different scenarios of climate change and management options?
What are the costs and benefits associated with the production of these goods and
services? How are they distributed between the different stakeholders?
CBA will allow the analysis of the evolution of marginal flow of costs and benefits by type of actors
(households, industry, agriculture, communities, etc.) according to alternative management of
woodlands.
32
Synergies C1/C2 in Algeria Assia Azzi
On the Chréa pilot site, goods and services selected for component 2 are:
Recreational aspect related to the Barbary macaque;
Production of Arbutus unedo;
Water purification
The main possibility of synergy is the use of the methodology for creating a vulnerability map to
predict possible changes on ecosystems and their capacity to provide goods and services assessed.
However, the main constraints are the following:
The pilot sites for component 1 and component 2 are not the same: Senalba (semi-
arid) for C1 and Chréa for C2 (from subhumid to humid);
The parameters used in planning methodology of the vulnerability map on the site of
Senalba may not be adapted to the case of Chréa.
In terms of opportunities, thanks to the activities of Component 1, two Algerian managers were
trained in the use of the methodology for producing vulnerability maps in the framework of the
forest administration (DGF). Moreover, the organization of a national workshop combining the 4
components for improved interaction and sharing of data is scheduled for September 2014.
Synergies C1/C2 in Lebanon Carla Jamous
Many data will be exchanged between the two components.
From component 1 to component 2:
Demographic and socio-economic data: population since 1995 (grazing pressure
estimation, harvest of timber and collection of medicinal and aromatic plants);
Forest data: forest inventory (distribution of tree species on the site), logging (cuts);
Biodiversity data: endangered species with a map of their distribution on site (Quercus
cerris and Juniperus drupacea)
From component 2 to component 1:
Data on grazing: forage units, types of forage
This information will be obtained through a field survey, to be conducted as soon as possible (from
June 2014 to be continued until spring 2015)
The economic analysis is expected from October/November 2014, following the collection of data.
33
Synergies C1/C2 in Morocco Abdelmohssin El Mokaddem
Ten goods and services were selected for the pilot site of the Maâmora forest.
Useful data that can be exchanged between components 1 and 2 are as follows: slope maps, maps on
the state of development, contribution of forests to the generation of income to household,
exchange of data on cork collection (for the period 1992-2009 and 2011-2014), crimes recorded,
WWF study on acorns, route maps.
Under component 2, two scenarios of evolution will be used for the cost-benefit analysis. They are
produced by component 1 (projection of the state of vulnerability), with the hypothesis 0 (reference:
normal development without planning/response) compared with an evolution with intervention.
Synergies C1/C2 in Turkey Özge Balkiz and Murat Turkes
Limitations for Component 2:
There is a lack of accurate data on the costs of managing the production of industrial
wood in the pilot sitel;
The function of protection of biodiversity is being evaluated at the moment;
The budgets of public and private organizations for the protection of biodiversity are
unknown;
It is difficult to assess the impact of increased recreational use of forest biodiversity
areas (lack of data on long-term monitoring of species);
There is a lack of reliable information on tourist visits on the site and on the net
income of organizations for recreational areas;
There is a lack of specific data on carbon sequestration (this could be solved using the
case of Lebanon);
There is a lack of detailed information on the number of hunters in the region and a
lack of long-term data on the activities of big game hunting.
Possible synergies between component 1 and component 2 are as follows:
Review of the literature on changes in annual growth of pine forests in the
Mediterranean (by the end of June 2014);
Review of the literature on the evolution of diseases and pests in pine forests in the
Mediterranean (end of June 2014);
Predictions of wildfire risk specifically for the pilot site (to be produced by mid-October
2014);
34
Results of the vulnerability analysis on the pilot site: changes in terms of climatic
conditions and species distribution (by mid-October 2014) and link with Component 2
(e.g., the most sensitive areas (with strongest changes) will be those where the
production of wood industry of Pinus brutia will decrease).
In terms of timetable: the data collection and vulnerability analysis will occur until October. From
that date, the cost-benefit analysis can be made.
Synthesis Valentina Garavaglia and Christophe Besacier
This common day allowed having time for working and brainstorming together. Data collections in
the two components are very complementary. The coordination and undertsnding of the project in
each country is quite good. With regard to component 1, this common day allowed focused better
focus of the activities on the overall results and objectives of the project, after two highly technical
workshops in April 2014 (GIZ) and May 2014 (VITO).
Such work in synergy is also planned for components 1 and 4, particularly during the debriefing
workshop of C1 at the end of 2014 in Antalya (Turkey).
35
Presentation of results of the first workshop for capacity building organized in collaboration with VITO Valentina Garavaglia (FAO - Silva Mediterranea)
This first one-week workshop, held in Mol (Belgium) from 19 to 23 May 2014, aimed to analyse the
"Remote sensing techniques and classification of changes in forest cover, land use and climatic
conditions of the pilot sites selected for Component 1".
A second week of training will be held in Rabat (Morocco) from 22 to 26 September 2014.
The objective of the work in cooperation with VITO is the production of maps on the evolution of
forest cover (and other wooded ands/other lands) using processing tools of free information (QGIS:
free mapping software; PkTools: a tool for image processing developed by VITO; Collect Earth:
Google Earth plugin developed by FAO for the systematic analysis of a sample of forest plots; OSGeo-
Live: system to try a wide variety of open source geospatial software without installing anything) and
free data (Landsat images in particular).
The approach is presented here below in a diagram:
Source: Fouad Mounir (Morocco).
The classification is the most complicated part of the study; it will be further discussed during the
workshop in Rabat. Before the workshop in Rabat, the data must be collected. At the end of the
workshop, the final maps and the results (analysis of changes in land use in particular) will be
produced.
36
The approach is also presented here below in more detail:
Source: VITO.
Notes:
Some national experts are not part of forest administrations. How can we sustain this expertise?
In the case of Lebanon, the methodology of VITO is useful for using free tools, but it is not very
appropriate given the small size of the pilot site.
A service contract was signed with VITO (assistance online or on-site assistance).
37
Presentation of GIZ activities on the vulnerability of ecosystems in Tunisia Abdelmajid Jemaï (GIZ)
The conceptual approach of the methodology used: multifactor spatial modelling is presented in the
following diagram:
Source: GIZ.
The methodology is explained in more detail in the report of the workshop of Solsona.
The results of the multifactor spatial modelling will allow assessing risks and estimate losses of
goods and services provided by forest ecosystems in pilot sites and it will answer to the following
questions:
What are the consequences of climate change impacts on woodlands and on their provision
of goods and services (2020 and 2050 projections)?
From which level we can speak of risk of loss for different tree species?
What is the shift of favourable conditions for the presence of a given ecosystem (e.g.
Suber)?
How to take into account future climate data (not just the current climate data) in future
management plans?
What are the concrete actions to implement in order to improve the adaptation of
woodlands to climate change?
Capacity building for different stakeholders (awareness, training) is necessary.
38
SESSION 5: Summary of exchanges in each pilot site to identify next steps and overcome difficulties to finalize the vulnerability analysis by the end of 2014
Concerning the multifactor spatial analysis:
- take two new scenarios of the 5th IPCC report: RCP4.5 (optimistic - old B1) and RCP8.5 (pessimistic -
old A2) timeline is from 2016 to 2035 and from 2046 to 2065
-take the models of WorldClim (http://worldclim.org/cmip5_30s): in Turkey, after tests by
meteorologists for choosing the best model, the one selected is HadGEM2-ES. Unless meteorologists
disagree, it will be removed from the pilot sites model.
- in WorldClim, timelines proposed are 2050 and 2070
- move the timeline to 2050 (which corresponds to an average of 2041-2060; this correspond more
or less to the second term proposed by the 5th IPCC report)
Concerning the 5th and last IPCC report (2013), see the following articles:
- Explanation of the choice of factors, advantages and disadvantages of the dataset available in the
pilot site
+ Production of a sheet with a synthetic presentation
- VITO and GIZ will be involved in drafting this final report for each pilot site
The last part of the report will focus on recommendations. These will be described by type of user
(managers, policy makers, researchers, etc.).
A poster could be prepared and submitted by each country especially during Fourth Mediterranean
Forest Week in Barcelona (Spain) in March 2015 and the World Forestry Congress in Durban (South
Africa) in September 2015.
41
It is also possible to publish articles in various newspapers (Unasylva, Forêt Méditerranéenne, etc.) or
scientific journals.
Concerning the last practical and financial aspects:
Morocco, Tunisia and Lebanon: LoA with local associations. The first instalment was paid (30% of the
total amount equal to € 15,000). Countries can make the request for the second instalment (50%
equal to € 25,000). For the second payment, there is no need for detailed documentation. The third
payment (20% equal to € 10,000) will be made once the work is completed. That thus requires a cash
advance at project completion by the partner organizations. For this third payment a detailed
justification of the total expenditure is needed.
For minor expenses, the allocation of expenses among the budget lines is flexible. A simple email to
inform FAO is enough. In case of major changes or questions, FAO/Silva Mediterranea should be
contacted.
Regarding the payment of national consultants recruited outside the LoA (through FAO
representations in Algiers, Ankara and Beirut), they must complete their working time declaration
form on the system to manage consultants and staff (Global Resource Management System - GRMS)
and send it for approval to FAO before the payment. The payment is not monthly, but it will be when
results are produced in accordance with Terms of Reference.
Algeria, Turkey: no contract with local associations FBA (Field Budget Authorization): Permanent
Representations of FAO in Algiers and Ankara are authorized to make expenditures.
Finally, Algeria and Tunisia, there is a limit for the release of foreign currency abroad which can cause
problems during missions outside the country show a certificate?
42
ANNEX 1: Agenda of the workshop in Tunis
Mid-term workshop
Project: Optimizing the production of goods and services of Mediterranean forests
in a context of global changes
Component 1: Production of data and development of tools to support decision and management of
vulnerable Mediterranean forest ecosystems affected by climate change and the ability of these forest
ecosystems to adapt to global change
2 - 5 June 2014
Hotel Belvédère Fourati, Tunis
Objectives
Main objectives of the mid-term workshop of Component 1 organized in Tunis on 2-5 June 2014 in
collaboration of the General Directorate of Forests (Government of Tunisia) are:
• Illustrate the first results produced by the partner countries on vulnerability assessment of
Mediterranean forest ecosystems to climate change in the pilot sites
• Illustrate the results produced by INRA Avignon on the state of the art of climate change impacts on
selected Mediterranean forest species in the pilot sites for the implementation of activities of the
project
• Discuss with the coordinators, experts, focal points and thematic referents of Component 2: (i) the
first results produced by Component 1, (ii) the data expected by end of 2014 by experts of
component 2 on the impacts of climate change on goods and services and (iii) future collaborations
and exchanges of results obtained by the end of 2014; iv) next steps for all activities scheduled for
Component 1
Participants
Focal points
Thematic experts
National experts
One expert from INRA Avignon
One expert from GIZ Tunisia
FAO – Silva Mediterranea: Moderator
Others: Representatives of the Tunisian General Directorate of Forests
Arrival of participants on 1/6/2014 – departure on 6/6/2014
Day 1 : 2/6/2014
9h00 – 9h45
Welcome (Tunisian General Directorate of Forests)
Round table to introduce participants (Valentina Garavaglia/FAO)
Objectives of the workshop (Valentina Garavaglia/FAO)
Review of activities scheduled for Component 1 (Valentina Garavaglia / FAO)
9h45 – 10h30 First results and program for 2014 : pilot site of Djelfa (Algeria)
Coffee break
11h00-11h45 First results and work plan 2014: pilot site of Jabal Moussa (Lebanon)
11h45-12h30 First results and work plan 2014: pilot site of Maamora (Morocco)
Lunch break
14h00 – 14h45 First results and work plan 2014: pilot site of Senalba (Tunisia)
14h45 – 15h30 First results and work plan 2014: pilot site of Düzlerçami (Turkey)
Coffee break
16h00-16h45
State of the art of climate change impacts on selected Mediterranean forest species in
the pilot sites of Component 1: first results and workplan for 2014 (Nadine Wazen,
INRA Avignon)
16h45-17h30
State of the art of activities/studies/projects on adaptation of Mediterranean forests
to climate change in areas similar to the pilot sites selected for Component
1 (Valentina Garavaglia/FAO)
17h30 – 18h00 Synthesis of the day and preparation of the joint day of Component 1 and 2
Day 2 : 3/6/2014 Joint day Component 1/Component 2
9h – 9h30 Welcoming and clarification of the objectives of Components 1 and 2 joint day
9h30 – 10h30 Presentation of ongoing activities, workplan 2014 and description of data produced in
the framework of component 1 (Valentina Garavaglia and Christophe Besacier/FAO)
Coffee break
11h00 – 12h00 Presentation of ongoing activities, workplan 2014 and description of data required for
the implementation of component 2 (Marion Duclerq/Plan Bleu)
12h00 - 13h00 Round table: initial exchange on potential synergies between Components 1 and 2
Lunch break
14h30 – 16h00 Exchange with countries to identify expected results and synergies to be developed
between component 1 and component 2
Coffee break
16h30 – 18h00 Outputs of countries, expectations and synergies to be developed between component
1 and component 2, and synthesis of the joint day (Christophe Besacier/ FAO)
Day 3 : 4/6/2014
9h00 – 10h30
Next steps to be taken for the compilation of the state of the art by INRA Avignon/
data to achieve the objectives of this study by the end of 2014 (Nadine Wazen/INRA
Avignon)
Coffee break
11h00 – 12h30
Presentation of the results achieved during the workshop organized in collaboration
with VITO Vision on Technology on "Production of cartographic material on the
evolution of forest cover, land uses and climatic conditions of the pilot sites"
(Valentina Garavaglia/ VITO)
Lunch break
14h00 – 14h45 Presentation of the activities implemented by GIZ Tunis on the activities performed to
assess the vulnerability of ecosystems in Tunisia (Abdelmajid Jemaï /GIZ)
14h45 – 15h30 Presentation of the results of the workshop organized in collaboration with GIZ Tunis
on ”Multifactorial spatial analysis of the vulnerability of pilot sites to climate change
selected for Component 1’’ (Kamel Tounsi/Ali Aloui/GIZ)
Coffee break
16h00 – 17h30
Next steps to be implemented after the first week of workshop organized in
collaboration with VITO on the production of cartographic material on the evolution of
forest cover, land uses and climatic conditions of the pilot sites
17h30 – 18h00 Presentation of MOSAICC, FAO Modelling System for Agricultural
Impactes of Climate Change (Renaud Colmant/FAO)
Day 4 : 5/06/2014
9h00 – 12h30
Exchange at each pilot-site level to identify next steps and overcome difficulties to
finalize vulnerability assessments by 2014
Lunch break
14h00-14h30 Next steps and workplan 2014: pilot site of Djelfa (Algeria)
14h30-15h00 Next steps and workplan 2014: pilot site of Maamora (Morocco)
15h00-15h30 Next steps and workplan 2014: pilot site of Jabal Moussa (Lebanon)
Coffee break
16h00-16h30 Next steps and workplan 2014: pilot site of Siliana (Tunisia)
16h30-17h30 Next steps and workplan 2014: pilot site of Düzlerçami (Turkey)
17h30-18h30 Summary of next activities and conclusion of the workshop
43
ANNEX 2: Methodology "Impacts of climate change on the distribution of native tree species in Lebanon potential projections by 2050"
the introduction and presentation of the objectives and methodology of the study conducted in
Lebanon. The full report is available - contact Miguel Ángel Navarrete Poyatos: [email protected].
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON:
POTENTIALITY PROJECTIONS TO 2050
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 2
CENTER FOR APPLIED RESEARCH IN AGROFORESTRY DEVELOPMENT (IDAF) - UNIVERSITY OF CÓRDOBATechnical Staff
Project Coordinator, IDAF Miguel Ángel Navarrete PoyatosForest Management Expert, IDAF Miguel Ángel Lara GómezGIS and Remote Sensing Expert, IDAF Jesús Trujillo ToroGeneral Manager and Director, IDAF Guillermo Palacios RodríguezHead of ERSAF Research Group, University of Córdoba Rafael María Navarro CerrilloPermanent Teacher and Researcher, Lebanese University Dr. Salim KattarProgramming Expert, University of Córdoba Joaquín Duque Lazo
The author’s views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government.
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON:
POTENTIALITY PROJECTIONS TO 2050
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 3
INDEX
1. INTRODUCTION
2. OBJECTIVES2.1 GENERAL OBJECTIVE2.2 SPECIFIC OBJECTIVES
11. ANNEX I. CLIMATE VARIABLES. ANALYSIS AND SELECTION11.1 CURRENT CLIMATE VARIABLES11.2 FUTURE CLIMATE VARIABLES11.3 VARIABLES SELECTION11.4 CURRENT AND FUTURE VARIABLES MAPS
12. ANNEX II. SPECIES DISTRIBUTION MODELS. DESCRIPTION AND EVALUATION12.1 STATISTICAL MODELS12.2 MODELS DESCRIPTION12.3 MODELS EVALUATION12.4 POTENTIALITY THRESHOLD
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 5
EXECUTIVE SUMMARY
Numerous studies conducted during the past decade show that climate change is most likely attributable to increased concentrations of anthropogenic greenhouse gases producing effects on climatic patterns. Under this new and changing situation, forest management and policies require new approaches that take into consideration the effects of climate change.
In recent years, governments, institutions, and NGOs in Lebanon are making huge efforts to carry out reforestation programs throughout the country. Taking into account the future potentiality of species currently used in reforestation by placing them under different climate change scenarios is a useful tool to understand which species will thrive in future conditions.
For this purpose, almost seven thousand points of presence and twelve thousand points of absence were identified in Lebanon. Ensemble projections for each species were obtained by adding climatic variables for current conditions and running models with Biomod2 R-package platform for future scenarios.
This study shows the Species Distribution Model for 2050. It takes into account twenty main native species commonly used in Lebanese reforestation, and places them under A2 and B1 IPCC scenarios. In addition, vulnerability classification of Lebanese territory was conducted in terms of species richness loss caused by climate change. This determines the critical areas to be restored and/or protected in terms of species diversity.
Finally, general guidelines for future management and measures for adaptation and mitigation to climate change are proposed for each species and case of potentiality status in the future. In terms of potential species richness loss, specific actions are set for the considered critical areas.
The generated maps and raster files will help the forest manager in decision-making regarding the priority areas for restoration or management and the potential species to be used considering the future effects of climate change.
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 6
1. INTRODUCTION
According to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC, 2007), climate change is unequivocal and most likely attributable to increased concentrations of anthropogenic greenhouse gases. The same report relates warming for the last three decades to the changes in many physical and biological systems at a global level. Forecasts suggest the effects will persist in the future.
This new and changing situation requires forest policies and management to be reoriented towards species adaptation to climate change. In Lebanon, where global models predict severe climate change and where numerous species grow in one of the most biologically diverse ecosystems, the Mediterranean forest, the knowledge of future climate behavior is an important management tool for the preservation of natural resources.
Therefore, priority should be to expand the knowledge of the vulnerability to and impacts of climate change on Lebanese species richness, so that adaptation can be designed and integrated into policy planning and management of biodiversity. It is with this knowledge that the actions for preservation in a world of changing climate will have the highest impact. However, it is important to keep in mind that the success of any adaptation measures will eventually be conditioned by the correction of the climate-altering causes.
Overall, this work analyzes the potential effects of climate change on key vegetal components in Lebanese ecosystems. It has been carried out with the best available knowledge on climate projections for the 21st century, taking into account the current distribution of target species.
This work aims at supporting the national reforestation program launched in 2010 by the United States Forest Service (USFS) office of International Programs (IP) through the support and funding of the United States Agency for International Development (USAID), implemented through the Lebanon Reforestation Initiative (LRI). This project also continues the collaborations between IDAF and LRI that started in 2012 regarding climate change and species richness.
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 7
2. OBJECTIVES
2.1 GENERAL OBJECTIVE
The identification of key areas to be restored and adapted to climate change effects, securing the level of biodiversity, setting a vulnerability classification of Lebanese territory in terms of species richness loss caused by climate change and development of ensemble projections under different climate change scenarios.
2.2 SPECIFIC OBJECTIVES
• Assessment of species richness in Lebanon determining the areas that could sustain more biodiversity than their current status;
• Identification of key areas that will lose more species richness due to climate change;
• Development of ensemble projections under different climate change scenarios (present till 2050) for each selected species, up to a total of around 20 species including endemic and indicator species if possible;
• Proposition of possible treatments and actions to be implemented, aiming at improving the adaptation of these species in key areas.
For this purpose and due to the scope of this project, twenty of the most important species for forest restoration in Lebanon have been assessed:
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 8
3. METHODOLOGY
3.1 SCOPE
The study area covers the entire territory of Lebanon, ranging up to 10,452 Km2. The spatial resolution of the results is 1 Km2, derived from the characteristics of the climatic raster files used in this study. Presence/absence and current/future potentiality of species has been developed for a grid of 1 x 1 km.
3.2 SCHEME OF WORK
In order to generate ensemble projections, models for the twenty selected species, climate variables, and points of presence/absence were gathered. After treating data, future models were run using a script designed with Biomod2 R-package under A2 and B1IPCC scenarios.
Current and future Species Distribution Models (SDMs) for species potentiality were obtained. A potentiality threshold was calculated for each species depending on whether it is a generalist species or a species with specific needs. The Model also takes into consideration whether the identified presence points are spread throughout the entire potential distribution range of the species or conversely, if the species grow in relict populations without covering all environmental variability. Using this threshold, classified maps of potential/non-potential areas were obtained.
Depending on the possible combinations of potentiality status in current and future scenarios, maps of management guidelines per species were proposed.
To assess vulnerability classification of the Lebanese territory, species richness maps were developed for current and future scenarios. The maps show the number of species that could potentially occur on each cell. Maps of potential loss/gain species richness were generated by subtracting future predictions from current status in both scenarios.
Once these maps were developed, critical areas were identified by making the arithmetic mean of both future scenarios and segmenting the resulting map into severe, important, and moderate classes by applying 75th and 90th percentiles. Areas where severe or important loss occurs are considered critical areas for adaptive management and restoration. Figure 1 shows the flowchart of this study.
Figure 1. Study Flowchart
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 9
4. CLIMATE VARIABLES
4.1 CURRENT CLIMATE VARIABLES
Current climate data were obtained from WorldClim (Hijmans et al., 2005), a set of global climate layers with a 30 arc-second spatial resolution generated through interpolation of real data from weather stations for the period 1950 - 2000.
Bioclimatic variables are derived from the monthly temperature and rainfall values in order to generate more biologically meaningful variables. These variables represent annual trends (e.g., mean annual temperature, annual precipitation), seasonality (e.g., annual range in temperature and precipitation), and extreme or limiting environmental factors (e.g., temperature of the coldest and warmest month, and precipitation of the wet and dry quarters). A total of 19 bioclimatic variables were used.
In addition to this, a new set of variables was generated clustering climatic information in seasonal periods, obtaining the minimum, maximum, and average temperature of each season and seasonal total precipitation.
The species distribution models were built upon the bioclimatic variables. These models were used to obtain the current potential surface.
4.2 FUTURE CLIMATE VARIABLES
To obtain more significant climatic variables, a mixed climate model was generated from predictions made by CCCMA-CGCM3 and ECHAM5 global models.
Once the model is defined, the conditions of the factors influencing climate evolution (i.e. the emission scenarios) should be determined. There are four families of scenarios defined by the IPCC (Intergovernmental Panel on Climate Change): A1, A2, B1 and B2. Each one is a combination of demographic, social, economic, technological, and environmental trends.
A2 and B1 scenarios were selected for this study, representing unfavorable conditions and less climate change impact respectively.
The horizon for the prediction of potential distribution is 2050.
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 10
4. CLIMATE VARIABLES
4.3 VARIABLES SELECTION
Collinearity analyses were run to eliminate repetitive variables with a strong correlation. The correlation coefficient and the variance inflation factor (VIF) were calculated. The analysis of collinearity was done within the full list of original variables. Variables with R2>0.90 and VIF>9 produced a poor estimation of the correlation coefficient due to collinearity and were deleted (Graham, 2003; Heikkinen et al., 2006).
Three common non collinear variables were selected:
For further information regarding climatic variables and statistical analysis, see Annex I.
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 11
5. SPECIES DISTRIBUTION MODELS (SDM)
The variety of techniques accessible to model species distribution can be classified in three groups, depending on the input data:
• Profile techniques which require species presence-only data (i.e. environmental hype-space inhabited by species methods as BIOCLIM, among others);
• Discriminative techniques which require species presence-absence data (i.e. General Linear Model (GLM), Maximum Entropy (MaxEnt), among others);
• Mix modeling approach which uses combined techniques (i.e. Biomod, among others).
To deal with model technique election, Biomod2 R-package (Thuiller et al., 2013) which includes ten SDMs techniques was selected for this assessment. Default settings of biomod2 (version 2.1.15) were used. Biomod2 R-package is a computer platform for developing ensemble SDMs, which works with presence-absence data.
In order to select the most accurate SDM and avoid drawbacks of different individual statistical models, only ensemble models obtained from the linear combination of the ten models evaluated by R-package biomod2 were used.
Models were run ten times per each species and only those that presented higher statistical parameters in all cases were selected. The final SDM was obtained by assembling selected models.
Obtained SDMs are graphical representations of the probability of the presence of a species in a particular geographical location. Each pixel represents the percentage of probability of presence for the species. In order to assist managers in decision making, these maps were reclassified indicating the probability threshold of the species.
This limit is calculated separately for each species from different statistical methods. The choice of appropriate statistical method depends on particular characteristics of the species distribution, the dispersion of its points of presence and absence, and their distribution on the generated model.
See Annex II for SDM detailed information.
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 12
6. IDENTIFICATION OF SPECIES PRESENCE / ABSENCE
The R-package requires information about current species distribution within the study area. For each species, points of presence were georeferenced. The number of points referenced was dependent on site accessibility, species populations, and security risks. Collection was spread throughout Lebanon and, absence points were evenly distributed in areas presenting unfavorable conditions for species to grow.
Spatial data resolution was respected while establishing presence/absence points by setting only one point per species inside each 1 x 1 Km2 cell; the same spatial resolution as the climatic data grid.
Points of presence were sampled on the field identifying as many populations possible for each species. To reach areas not accessible by vehicle, field trips were designed covering large areas through dense forests and riversides. A large amount of points were obtained with GPS devices. After processing and downloading points to a GIS platform, cells of the 1 x 1 Km2 grid were analyzed to ensure that only one point per species was located inside each cell. By using recent orthophotos of Lebanon, homogeneous forest areas were located and additional points of major species were set nearby sampled areas.
In areas not accessible due to the political situation, historical records and analysis of high-resolution images such as Google Earth or Bing Maps were studied. Interviews were conducted with botanical experts in Lebanese flora. Pictures taken in situ by locals were also analyzed to identify species occurring in these remote areas.
A total of 6,767 points of presence and 11,851 points of absence were set. Figure 2 shows the main location where points of presence were identified.
Figure 2. Principal Locations for Presence Points
CLIMATE CHANGE IMPACTS ON NATIVE TREE SPECIES DISTRIBUTION IN LEBANON: POTENTIALITY PROJECTIONS TO 2050 13
7. SPECIES POTENTIALITY
7.1 OBTAINED RASTER FILES
Attached to this document are the raster files. Metadata can be delivered on computer format. The information found in these files is listed below:
• Metadata file identifier • Language • Character set • Contact • Metadata date stamp • Metadata standard name • Metadata standard version • Spatial representation information • Identification information • Distribution information • Application schema information
7.2 OBTAINED SPECIES DISTRIBUTION MODEL
The following maps show the current potentiality for the studied species versus the expected future potentiality in 2050.
The map legend shows the points of presence/absence identified for the species, the potential threshold and the total potential surface for present situation under both B1 and A2 scenarios.
These maps are classified into six different potentiality classes expressed in percent. In some cases, potentiality threshold is located within the range of values of one class and the limits of potential areas are not defined clearly. For classified maps showing potential or non-potential areas for the species without discerning between potentiality ranges, see Annex III.