EFIMED Advanced course on MODELLING MEDITERRANEAN FOREST STAND DYNAMICS FOR FOREST MANAGEMENT Mediterranean forest management and planning: the need for simulation models MARC PALAHI Head of EFIMED Office
Mar 31, 2015
EFIMED Advanced course onMODELLING MEDITERRANEAN FOREST STAND
DYNAMICS FOR FOREST MANAGEMENT
Mediterranean forest management and planning: the need for simulation models
MARC PALAHIHead of EFIMED Office
20.8.20042
Contents
1 Some features of Mediterranean forests
2 Some concepts on forest planning
3 Simulation: a key step
20.8.20043
Features of Mediterranean forests
Long history of manipulation by man
Many types of natural vegetation and high biological diversity
Relevance of their protective, social and ecological functions
versus the productive ones (externalities)
Fragility, instability, over-exploitation (south) & fires (north)
20.8.20044
The natural vegetation
The climate and the altitude factor makes possible various vegetation zones:
-Thermo-Thermo-, MesoMeso-: Q. ilex, Q. suber, P. halepensis, P. brutia
-Supra-: Q. robur, Fraxinus spp., P. nigra-Montane- and Oro-: Cedrus, Fagus sylvatica, P.
sylvestris
Agro-silvo-pastoral systems: dehesa or montado
20.8.20045
High biological diversity
The Mediterranean area harbours 25000 plant species (50 %
endemic) whereas in central and northern Europe (an area 4
time greater) 6000 plant species can be found
In forest tree species: 100 vs. 30, respectively
20.8.20046
Many non-wood forest outputs
highly demanded by the society• non-wood products: cork, grazing, resin, mushrooms,
aromatic plants, honey, fruits (pinecones and acorns), truffles, game,
• services and externalities: soil protection, flood and avalanche prevention, landscape quality, nature conservation, recreation possibilities, micro-climate regulation
20.8.20047
Forest management planningComplex problem because of multiple
competing/complementary objectives/products
NEEDS
New models, techniques and tools to support decision
making in forestry
20.8.20048
Multi-objective forest planning
1 Features of Mediterranean forests
2 Some concepts on forest planning
3 Simulation: a key step
20.8.20049
Forest Planning?
Planning
• finds the optimal way to use forest resources
• maximises the production of goods and services
• those goods and services considered which are important to the forest owner/society (goal based)
• always utility maximisation & optimisation
20.8.200410
Planning, not an easy task…
Many goals to address the multiple functions
Many parties (locals, government, ecologists)
Long time horizons (as compare to agri)
Risk and uncertainty
Many alternatives
20.8.200411
Framework of modern planning- A quantitative approach
Decision maker Forest ecosystem
Inventory dataPreferences Models
Objectives and constraints
Information about alternatives
Comparisons
Decision
SIMULATION
OPTIMIZATION
Growth models
Fire risk models
Habitat models
Mushroom models
20.8.200412
Simulation: a key step in planning
Decision Support
Planning
Simulation
Data Management
Measurement
Forest Ecosystem
Decision maker Forest ecosystem
Inventory dataPreferences Models
Objectives and constraints
Information about alternatives
Optimizations
Plan
Simulation
20.8.200413
Stand development
GrowthMortality Ingrowth
20.8.200414
Forest stand development affected byRegenerationGrowth of trees MortalityHuman interventions
Models should be able to predict these processes which are
affected by factors like
• Productive capacity of an area – site quality
• Degree to which the site is occupied – density/competition
• Point in time in stand development - age
20.8.200415
Modeling forest development is needed to
provide tools that enable foresters to compare alternative
silvicultural treatments
predict the economic returns of a management schedule but also to
produce information about the dynamic change of less tangible
attributes of forests
generate silvicultural instructions for different species, sites and
management types
0
100
200
300
400
500
600
700
800
900
0 20 40 60 80 100 120 140
Age (years)
Vo
lum
e (
m3/h
a)
SI-17
SI-24
SI-30
Simulation of silvicultural alternatives
0
10
20
30
40
50
60
40 60 80 100 120 140
Stand age (years)
Ba
sa
l are
a (
m2 h
a-1
)
20.8.200416
Simulation modelscomprise of:
• a series of mathematical equations
• numerical values embedded in those equations
• the logic necessary to link these equations
• the computer code required to implement the model on a computer
Mathematically, e.g.;
Dbh-Increment = a + b dbh + c Site + d Basal area
Height = a + b dbh
Volume = a + b Height + c dbh2
….
Modelling
20.8.200417
SimulationTreatment schedules for stands
• Example:
1. Do nothing
2. Thinning
3. Clear-felling / selective cutting
Purpose: produce information for planning
• Stand development
• Harvested timber/firewood volume
• Costs and incomes
• Biodiversity indices
• Non-wood forest products
Results in a Decision Space
= Combinations of stands’ treatment schedules
20.8.200418
Simulation programSimulation program combines
• inventory data
• models
• rules for producing alternatives (”instructions”)
Models used
• models on stand dynamics
• regeneration
• growth
• mortality
• models on allometric relationships
• height = f(diameter)
• volume = f(diameter, height)
• other models (fire risk, mushroom, habitat models)
20.8.200419
Multi-objective Planning
Write a planning model using
• information from simulations by using models
• information on preferences -> objectives
• Planning model writer
Solve the model using
• mathematical programming
• heuristics
20.8.200420
Decision support systems
• Computer system which supports rather than replaces the
decision maker
• User-friendly interface
• Planning system (data base, simulation and optimisation)
augmented with e.g.
• Comparison tools
• Visualisation tool
20.8.200421
20.8.200422
20.8.200423
EFIMED Annual meeting 26-27 of october 2007
University of Valladolid, Palencia, Spain
Modelling the production of wild mushrooms
in Scots pine (Pinus sylvestris L.) forests in the Central Pyrenees
José Antonio Bonet , Timo Pukkala, Christine Fischer, Marc Palahí, Juan Martínez de Aragón, Carlos Colinas
20.8.200424
New forestry context
Recent socio-economic changes have accentuated the
multifunctionality of forest ecosystems
- economic development/ increasing living standards
- time for leisure and environmental awareness
- urbanization of society/depopulation of rural areas
- lack of man power and profitability (high costs, prices-globalization)
From Productive functions to Environmental and Social functions, which
need to be addressed in forest management planning (biodiversity,
recreation, scenic beauty, non-timber products, etc)
20.8.200425
Importance of mushroomsContrary to timber, non-timber products
have maintained their prize
Market demand has increased
Annual revenue from 478 metric tons of L. deliciosus sold in the
central Barcelona market (Mercabarna) is estimated at 1.5 - 2 million €.
Mushroom picking is a major recreational activity
valuation studies estimate that Catalans are willing to pay an
average of 6 € year-1 to be able to pick mushrooms
Pinus sylvestris forests in Catalonia can produce 60 kg ha-1 of edible
mushrooms
20.8.200426
Managing forests for mushroom production
The social and economic importance of mushroom picking requires
mushroom production to be an explicit management objective in forest
planning
Quantitative scientifically based forest planning requires models to
predict the yield of mushroom according to forest stand characteristics
and management practices
No such predictive models
were found in the literature
for mushrooms
20.8.200427
Aim of the study
To develop empirical models for predicting the production of wild
mushrooms in Scots pine (Pinus sylvestris L.) forests in the Central
Pyrenees based on mushroom production data from three
consecutive years.
20.8.200428
Mushroom data
In 1995, 36 plots of 10 x 10 meters were established in Pinus sylvestris
plantations of the Central Pyrenees to evaluate the productivity and
diversity of ectomycorrhizal and edible fungi in this forest community.
The plots were sampled at 1-week intervals from September through
November during the 1995, 1996 and 1997 autumn seasons
We used the following groupings in the model: all species, the edible
species, the marketed edible species, and the marketed edible
Lactarius species.
20.8.200429
Forest data
24 plots inventoried in 2006 to measure site and growing stock variables
(other plots had been cut or significantly transformed through management)
Plots (area varying between 0.04-0.16 ha) were established so that at least
100 trees with dbh> 7.5 cm were within the plot.
Dbh and the growth for the last ten years were measured for all trees and tree
heights, tree age and bark thicknesses were recorded for a sample of at least
20 trees per plot.
Variable Mean Standard deviation Minimum Maximum
Stand variables
T (yr) 27.9 12.4 10.4 55.3
Hdom (m) 12.3 4.4 3.3 20.0
G (m2 ha-1) 20.6 13.6 1.0 54.8
Ntrees (trees ha-1) 1171.7 392.3 717.2 2196.3
Dm (cm) 17.2 7.0 4.9 34.2
SI (m) 22.3 2.9 13.3 27.5
Elevation (m) 1238.8 220.2 846.0 1528.0
Aspect (degrees) 179.9 130.4 4.0 356.0
Slope (%) 24.1 7.2 7.0 38.0
Mushroom productions
Total (kg ha-1) 123.7 135.2 0.2 466.6
Edible (kg ha-1) 63.0 75.0 0.2 283.4
Marketed (kg ha-1) 25.6 39.1 0.2 153.4
Lactarius (kg ha-1) 7.9 21.6 0.0 104.5
20.8.200430
Modelling mushroom production
Mushroom production depends very much on weather conditions but
also on the forest site and growing stock characteristics
However, in forest planning, variables that can be known trough regular
inventories and simulation tools and that can be influenced through
forest management need to be used
Weather conditions cannot accurately be predicted beyond a few weeks
20.8.200431
Modelling approach
The predictors were chosen from stand and site variables as well as their
transformations (Age, Site index, Hdom, Basal area, N, ELE, SLO, AS, …)
Due to the hierarchical structure of the data, mushroom measurements of
the same year were correlated observations as were the measurements on
the same plot, the generalised least squares (GLS) technique was applied
to fit mixed linear models
ijjinij xxxfy ),...,,( 21
Models for the total production, edible species, marketed species, and individual species or species groups were fitted
20.8.200432
Results
The regression analyses showed that stand basal area, elevation, aspect and slope were the most significant predictors:Total productionln(yij) = 0.981 +2.483ln(G) -0.128G +0.934cos(Asp) -0.0135Slo1.5 + ui + uj + eijEdible mushroomsln(yij) = -4.329 +1.966ln(G) -0.118G +0.636cos(Asp) +0.00331Alt + ui + uj + eij Marketed mushroomsln(yij) = -6.236 +1.246ln(G) -0.0599G +0.00459Alt + ui + uj + eijMarketed Lactariusln(yij) = -0.192 +1.016ln(G) -0.106G +1.489cos(Asp) -0.0151Slo1.5 + ui + uj + eij
20.8.200433
ResultsElevation 1240m, Slope 24%, East
0
20
40
60
80
100
120
0 10 20 30 40
Basal area, m2ha-1
Fre
sh w
eigh
t, k
g ha
-1
Total
Edible
Marketed
Lactarius
Altitude 1240m, Slope 24%
0
50
100
150
200
250
300
350
400
0 10 20 30 40
Basal area, m2ha-1
Tot
al f
resh
wei
ght,
kg
ha-1
North
EastWest
South
20.8.200434
Results
050
100150200250300350400450500
0 20 40 60
Basal area, m2ha-1
To
tal y
ield
, kg
ha
-1
0
100
200
300
400
500
600
700
0 10 20 30 40
Slope,%
To
tal y
ield
, kg
ha
-1
0
100
200
300
400
500
-1 -0.5 0 0.5 1
CosAspect ("Northness")
To
tal y
ield
, kg
ha
-1
20.8.200435
ResultsProductions were greatest when stand basal area was approximately 20
m2 ha-1.
Increasing elevation and northern aspect and decreasing slope
increased total mushroom production, edible and marketed
Marketed Lactarius spp., the most important group collected in the region,
showed similar relationships.
The annual variation in mushroom production correlated with autumn
rainfall.
20.8.200436
Discussion It seems that highest mushroom production coincides with the peak in forest
volume growth Previous studies shows that mushroom production correlates with growth
and photosynthetic rate of host trees
• Flux of current photosynthates is critical for soil respiration and
ectomycorrhizal sporocarp production
Since stand basal area is correlated with site conditions (soil quality, water
availability, humidity, etc.), and other variables like age, volume, etc.
estimating the effects of growing stock variables requires more plot
measurements or a population in which stand variables are less correlated
20.8.200437
Discussion Elevation, aspect and slope in the Prepyrenees range reflect water
availability and soil quality which clearly affects mushroom production Stands near canopy closure with vigorous growth rates located at high
elevations, in northern aspects and with low slopes seems to be optimal sites
for mushroom production in Scots pine forests of the Spanish pre-Pyrenees. Such models can be used to optimized forest stand management for
mushroom and timber productionDespite of the limitations of our data in number of measurements and plots
the results of the study are encouraging because they demonstrate that
mushroom production are related to stand characteristics that can be
influenced by silvicultural interventions
20.8.200438
Next step is to collect a larger data set including more variability in
growing stock characteristics as well as other tree species.
Collecting large quantities of empirical data over several years is required
because there are multiple factors responsible for high temporal variation
in mushroom productions.
The effect of silvicultural treatments needs to be studied
Climate change should be considered as it will affect both forest growth
(and composition) and weather conditions = mushroom production
Final remarks
20.8.200439
BONET, J.A. PUKKALA, T.; FISCHER, C.R.; PALAHÍ, M.; MARTÍNEZ DE ARAGON, J. i COLINAS, C. 2007. “Empirical models for predicting the yield of wild mushrooms in Scots pine forests in the Central Pyrenees”. Annals of Forest Sciences (in press).
MARTÍNEZ DE ARAGÓN, J.; BONET, J.A.; FISCHER, C.R. i COLINAS, C. 2007. “Productivity and richness of ectomycorrhizal and edible forest fungi in three pine forests of the pre-Pyrenees, Spain: Development of predictive models as a basis for forest management of the mycologic resource”. Forest, Ecology & Management (doi:10.1016/j.foreco.2007.06.040).
BONET, J.A.; FISCHER, C.R. y COLINAS, C., 2004. “The relationship between orientation and forest age on the production of sporocarps of ectomycorrhizal fungi in Pinus sylvestris forests of the Central Pyrenees”. Forest, Ecology and Management, 203: 157-175.
REFERENCES
20.8.200440
Dibujo transición corta setasDibujo transición corta setas
THANK YOU!!