fc - 7.1.2009

Annabelle Amm (INRA URFM), Daniel Auclair (INRA AMAP), Francois de Coligny (INRA AMAP), Véronique Cucchi (Cirad), Jean Dauzat (Cirad AMAP), Hendrik Davi (INRA URFM), Christine Deleuze (ONF), Nicolas Donès (INRA PIAF), Sébastien Griffon (Cirad AMAP), Degi Harja Asmara (ICRAF), Marilyn Laurans (Cirad AMAP), Isabelle Lecomte (INRA SYSTEM), Cécile Madelaine (IRD AMAP), Mohamed Najar (FCBA), Marie-Ange Ngo Bieng (Cirad), Raphael Pélissier (IRD AMAP), Thomas Perot (Cemagref), Michael Poissonet (FCBA), Eric Rigolot (INRA URFM), Vivien Rossi (Cirad), Lucile Soler (IRD AMAP), Grégoire Talbot (INRA SYSTEM), Oana Vigy (INRA URFM)

Excused: Olivier Pain (FCBA), Robert Schneider (UQAM - Québec)

pdf presentation

As an introduction, reminding of the project objectives and its organisation.

The capsis training politics and the 2008 session for 10 people.

The actives projects in 2007-2008:

• Prunus
• Oakpine2
• Sylvogène
• Regix
• FireParadox
• Mustard for the Deduction ANR project
• Migration
• Dynaclim
• JackPine
• Stretch

The main meetings in 2007-2008 where Capsis was presented.

The publications arround Capsis in the PMA06 conference proceedings.

The transfert actions 2006-2008.

The perspectives: mainly the recruitment of another INRA engineer to work on the project to work better and ofefr a better support to the modellers and end-users.

ppt presentation pdf poster

Context: Fire behaviour modelling, post fire ecological effects assessment, fire regime integration in forests dynamics.

A 3D model for fire propagation: Firetec.

Fuel description and modelling.

Fuel manager: user friendly software for input files generation of fuel complexes for fire behaviour models.

The FireParadox scene editor.

The EuroForestFuel database management.

The FireParadox pattern editor.

The export towards the fire model.

Perspectives: forest dynamics, management and climate change.

ppt presentation

Short description of the Short Rotation Coppice (SRC) of polar:

• field preparation
• installation during first year
• first rotation harvest at year 7
• second rotation with and without thinning at age 2

A stand model per species

• mean height
• mortality
• basal area
• biomass

Implementation for Poplar and eucalyptus

Beginning from a bare field or from an existing stand.

Working processes (harvesting, forwarding, logging…) with various objects (standing trees, logs, residues…)

Economic part: costs (cultural itinerary, harvesting, others…) and incomes (sales, others…)

Economic global billbook and summary, sensitivity analyses.

Evolutive software within Capsis, possible to add species and working processes.

pdf presentation

Mixed forests are 50% of the french forests (IFN 2008):

• how will they adapt to climate change ?
• interesting for biodiversity conservation (better than pure stands) ?
• How to manage them ?

Context: Quercus petrae, Pinus sylvestris (Cemagref / ONF project).

Develop a dynamics model: growth (this presentation), spatial structure, regeneration, mortality.

Sampling on 9 plots, competition indices in a 10m radius disk, CIpines and CIoaks competitors arround the target tree.

The Oakpine1 model in Capsis has a strong constraint: you need a map of the stand, rare.

Generate a virtual stand: Marie-Ange Ngo Bieng and Francois Goreaud (Cemagref) developped in 2007 a model to generate spatial structures with realistic spatial properties. Difficulty: define the spatial type of the stand.

Other approach: use the neighbourhood information of the trees, e.g. for 1 oak, n oak neighbours and m pine neighbours. Distribution of the numbers oak / pine (4 distributions). Rebuild neighbourhood for each tree gives a stand after simulation.

Oakpine2 is a distance-independent tree model, reconstruction of neighbourhoods. Computation for the competition indices of the growth model.

Perspectives: compare approaches Oakpine1 and 2 with real and simulated data.

ppt presentation

Lumber industry in Quebec: mainly spruce, pine and fir.

Jack pine is the 2nd most important reforestation species in the province.

Reasons for changing spacing: lower costs, larger trees, lower mortality. But what about quality ?

Need for a tool to predict the effect of different scenarios on the growth and wood quality of Jack pine.

Choice to use PipeQual by Makela and Makinen (2003). PipeQual is complex, needs a simulation software.

PipeQual was coded in Fortran and Matlab. Capsis was chosen to transfer the project to partners (more user friendly).

PipeQual structure and integration into Capsis.

Traditional Growth and Yield outputs, but also tree level details with tree taper and nodes / inter nodes biomass.

Work in progress and future developments.

Species are distributed in stable areas when climate is stable, but when climate changes, adaptation and migration. If change is fast, what capabilities for migration ? Migration is a change of distibution zones.

Importance of dispersion. Bioclimatic zones can move of several hundred of kilometers. Will forests be able to migrate ?

Focus on Silver fir (Abies alba) and Beech Fagus sylvatica).

Simulation to year 2100 shows the climatic zones of these species reduce. Evaluation of the risks for the forest trees.

Specific context: Mont Ventoux, reforested by pine, recolonized by fir and beech.

Seeds disperse with the wind and fall directly on the ground. If late fructification or short dispersion distance, possible problems for survival.

Methodology: analysis of all the processes involved in the forest dynamics, calibrated with climatic variables.

Capsis contains already several related models: Ventoux, VentouG, Dynaclim, Prunus.

Scale: 100 years, several kilometers, type of model: individual based model when short distance of dispersal, spatialized cohorts (can move) when densities are high. Climate files.

Cells on the ground for seed rain precise determination, seedlings in the cells (not spatialized).

Factors affecting the setting of the seedlings.

Bigger cells for climate.

Problem: how a cohort can disperse ? A cohort is a group of trees with same diameter, shape, position, density of trees and creation date. Each cohort has a fructification depending on diameters. Analysis with R, looking at the seed rain for various densities.

In Migration, circular cohorts (may change). Problem: aggregation / disaggregation of the cohorts.

pdf presentation (converted from ppt and compacted)

Forests dynamics is impacted by climate change, also acting on vegetation functionning, resulting in CO2 increase.

Model forest carbon and water cycle, estimate carbon (respiration) and water fluxes (rain, drainage, evaporation).

Castanea is a process based model developped in Orsay, compartments average tree model (canopy, trunk, soil…).

1. Leaf carbon budget (nitrogen contents) respiration, transpiration…
2. Integrate this budget into the canopy

Connected to a soil water model.

To parametrize: many mesurements to calibrate all the processes. The effect of the global change can be very different when changing geographic regions or ecosystems.

Coupling functionnal and dynamics models.

Mont Ventoux: the pine ecosystem will be replaced by a fir and beech ecosystem.

Ventoux models by Philippe Dreyfus (INRA URFM). Potential reductor growth, semi spatialized in square ground cells, mortality with probabilities per species, regeneration, see Annabelle Amm's work.

Change the portential of the dendrometric model with climate entries.

Coupling: choose a common spatialized representation (trees, cohorts, average trees).

Hybrid approach:

• in the tree level cells, model for the individual trees dynamics
• in the flux level cells, process based model: available carbon for growth, conditions the equations of VentouG

Dynaclim mixes Castanea and VentouG.

ppt presentation

1. Lucile Soler

Main interests:

• crown shape plasticity (plasticity is a major orientation of AMAP),
• increase the versatility / genericity of the model,
• modify the processes,
• Link with biomecanics within Capsis (later),
• transfer.

Stretch is a spatialy explicit, individual based, multi species, tri dimensional model with a yearly simulation loop. Crowns are optional:

• growth: stem volume with a Chapman Richards function
• crown growth depends on the stem growth, descriptes by virtual vectors branches and polygons
• optional use of Lieberman model with an index of closure of the canopy to characterise the light environment
• species dependent light growth reducer
• primary mortality (vigour and senescence)

Half stage, still work under progress.

2. Degi Harja Asmara

Focus on the light models SLIM and MMR.

SLIM: Simple Light Interception Model, based on hemispherical pictures, crown polygons see more or less light.

Validation: comparison on the plot with simulated hemispherical pictures.

STRETCH: Shape Transformation REsponse of Trees in Crowded Habitat.

MMR (Mir Musc Radbal) is a module in the Archimed platform by Jean Dauzat (Cirad AMAP), models direct and scattered light.

• Mir: incident radiation
• Misc: multiple scattering
• Radbal: radiation balance

Possible to have a light map on the ground.

MMR will be usable in Capsis from other forest dynamics modules.