1 / 43

Critical loads in France : links between dynamic modelling and biodiversity Anne Probst 1

Critical loads in France : links between dynamic modelling and biodiversity Anne Probst 1. with Estelle Bortoluzzi 1 , Arnaud Mansat 1 , Salim Belyazid 2. 1 CNRS, EcoLab, UMR CNRS-University Toulouse. 2 Belyazid Consulting & Communication AB. Context: The Nitrogen cycle sketch.

luna
Télécharger la présentation

Critical loads in France : links between dynamic modelling and biodiversity Anne Probst 1

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Critical loads in France : links between dynamicmodelling and biodiversityAnne Probst1 • with Estelle Bortoluzzi1, Arnaud Mansat1, • Salim Belyazid2 1CNRS, EcoLab, UMR CNRS-University Toulouse 2 Belyazid Consulting & Communication AB

  2. Context: The Nitrogen cycle sketch

  3. Change in river water nitrate concentration Change in nitrate-N concentration over 20 years River Derwent UK Cresser et al., 2008

  4. Change in soil biodiversity Effects of N treatment on collembola population density and biodiversity of alpine forest soils : 25 years of 12 kg.ha-1.a-1 atmospheric N deposition Xu et al., Env. Poll., 2009

  5. Change in vegetation… Eutrophisation/acidification : Galeopsis tetrahit in eastern France (Haut-Rhin) +13% 1999 1988

  6. Kt Context : Atmospheric Pollution • France pollutant emissions • SO2 decreases significantly by 1980 • NOx decreases but to a less extent by 1990

  7. Evolution of S atmospheric deposition Sulfur deposition has decreased in all sites Annual S atmospheric bulk deposition from 1993 to 2008 for 6 RENECOFOR sites (Kg. ha-1.yr-1)

  8. Evolution of N atmospheric deposition NH4 NO3 N deposition almost stable -generally slight decrease in NH4 -NO3 stable (See Poster by Pascaud et al. for details) Annual N atmospheric bulk deposition from 1993 to 2008 for 6 RENECOFOR (Kg. ha-1.yr-1) (NH4 : orange and NO3 : yellow).

  9. Probst et al., 2003 Critical load for Nutrient N Nutrient nitrogen for french forest ecosystems Europe, as a whole sensitive Sensitive areas Areas of low sensitivity Slootweg et al., 2007

  10. Critical load for Nitrogen Empirical modelling • Biogeochemical modelling CL(N)nut-CL(N)emp > 400 eq.ha-1.a-1 for 50 % ecosystems • Integrating interactions between N & plant biodiversity • Plant species as bioindicators • N impact on plant diversity Need of a third modelling approach CL=Ni+Nu+Nde+Nle(acc) Oak forest→ CL=17 kgN.ha-1.a-1 Alpine grassland → CL=7 kgN.ha-1.a-1 ... Flux : Ni immobilised OM, Nu uptake by vegetation, Nde denitrification Nle(acc) leached Probst et al. 2007

  11. Coupling biogeochemistry-ecology • Critical load calculation must take into account an approach combining biogeochemistry and ecology DOSE Biogeochem Models Biogeochemical Response Ecological Models Ecological Response Atmospheric deposition Soil Trophic characteristics Vegetation Biodiversity Approach and hypothesis ForSAFE-Veg (Sverdrup et al., 2007) Determine: • - Objectives for vegetation biodiversity protection • trophic critical limits : ecological models • critical loads : biogeochemical models

  12. Work progression • Three steps : • ForSAFE : • Site scale. • Modelling biogeochemical cycles. • Detailed soil processes. • ForSAFE-Veg : • regional scale. • Soil vegetation model chain. • Test the feasability and requirements needed to establish Critical loads based on vegetation changes. • VSD+ : • model well adapted to national scale, less input data . • European integration with the 2010 CFD. • VSD+ _Veg…next step

  13. VSD+ • NFC work • First trial of the model • Call For Data 2010

  14. VSD+ • Trial of the new VSD+ model (Bonten et al, 2009) : • 4 sites well documented with : • gradient in C/N ratio • Gradient in N deposition • Different kinds of vegetation. *: RENECOFOR data (ICP forests level II) Brethes et al (1997) ** : 2006 EMEP deposition.

  15. VSD+ inputs • Input parameters for the 4 studied sites: • Soil parameters • a single layer, enough data on a large number of sites. • Weathering • Profile model and total soil analysis. • Deposition • Biomass cycling of nutrients • Vegetation (type, age and growth rate) • No effect observed with different harvesting scenarii. • Mineralization, immobilisation • Default values used • Nitrification and denitrification • Default values used.

  16. VSD+ results • S and N deposition gradation between sites. • Cpool trend well predicted. • Npool simulated overestimated (impact of the artificial N initial deposition). • pH trends for PM40c and SP57 seem to reach a restored level.

  17. Comparison CL / EMEP deposition Vosges : slightly exceeded Landes : exceedance 1955 to 2000 Massif Central : never exceeded Moncoulon et al., WASP, 2007

  18. VSD+ input limitations • VSD+ message restriction : « Too little N available » • Influence on simulation results • Initial parameter calibration : overestimation • simulated C/N ratio • C pool Influence on simulation results: to be checked at the training session!

  19. ForSAFE-Veg • biogeochemical-vegetation model chain.

  20. ForSAFE-Veg • ForSAFE : more detailed, allows a better understanding of soil processes (Wallman et al., 2005; Belyazid, 2006) • Soil data on different layers • Hydric soil capacity • Climate data in monthly resolution (T°min max, mean; Pmm; Light) • Validation output data (soil solutions, biomass, nutrient content in needles, forest vegetation relevés). • Useful for testing the model chain with Veg. • A lot of input data needed • Not adapted at a national scale. • After validation : foreward to VSD+-Veg!

  21. The Veg model • Extension of the list of species : • Relevant species (common and/or characteristic) • Representing the various French forest ecosystems. • Helpul for a large part • of european ecosystems. Expert meeting (french-swedish collaboration): literature and database compilation Source : Bundesamt fur Naturschutz (BfN), 2003. Map of natural vegetation of Europe. Increasing number of species from 97 to 233

  22. Vegetation map for France • Towards a biogeochemical-phytoecological model • Phytoecological model C/N Occurrence Probability pH Databases EcoPlant CNRS, LEFORB, IMEP, INRA • 95 VMU + urban and water areas • ≈ succession types • 157 vegetation type units +Sophy + IFN + RENECOFOR + vegetation for France

  23. The Veg model • List of Veg-parameters : • Nitrogen (promoting and retarding classes) • pH (limit) • Water (promotion and saturation) • Light (saturation) • Temperature envelope • Height (effective shading for ground vegetation) • Rooting depths classes • Delay time • Browsing (food palatable classification). In the Veg-table for 233 species Extract of Fr-Veg table.

  24. The Veg model Veg combines the drivers…

  25. The ForSAFE-Veg model chain …provided by the biogeochemical platform ForSAFE (Belyazid, 2006; Sverdrup et al., 2007)

  26. ForSAFE • Trial of ForSafe (Wallman et al., 2005) : • 4 sites well documented with ENOUGH DATA! • Low gradient in C/N ratio • Gradient in N deposition • Different kinds of vegetation. • Gradient in CEC *: RENECOFOR data (ICP forests level II) Brethes et al (1997) ** : 2006 EMEP deposition.

  27. Forsafe inputs • Differences between throughfall EMEP data and RENECOFOR data measured : • Scale down of the EMEP data (needed from 1880). • Deposition scenarii until 2100 : • BGD : Background (like 1880 deposition) • MFR : Maximal Feasible Reduction (from 2010) • CLE : Current Legislation European (from 2010) • NOC : No Control (since the highest values of the 1980s). • Climate scenarii until 2100 : • First runs without taking into account climate change : a perspective!

  28. ForSAFE evaluation • First blind run • Soil chemistry data: outputs Modelled / measured data time series • Soil solution 20 cm site SP57 • Same pattern for others. • Cl- concentration: • Conservative : drainage calibration • Some overestimations • SO42- concentration: • overestimations (same Cl) • EMEP Deposition ? A Cl (µeq.l-1.month-1) Years B SO4 (µeq.l-1.month-1) Years Soil solution (-20cm) comparison between ForSafe monthlymodelled data and measured Renecofor data from 1993 to 2009 for Cl (A) and SO42- (B).

  29. ForSAFE evaluation • Soil chemistry data : outputs • Modelled / measured data time series • Na+ concentration : • Na: weathering calibration. • Underestimations: hydrology? • BC concentration : • Underestimation : usual in ForSAFE runs….. • But here : site dependant • To be solved: role of branches? role of deposition A Na (µeq.l-1.month-1) Years B BC (µeq.l-1.month-1) Years Soil solution (-20cm) comparison between ForSafe monthlymodelled data and measured Renecofor data from 1993 to 2009 for Na (A) and BC (B).

  30. EPC 08 Soil chemistry data outputs Modelled / measured data time series • Na concentration: • Good agreement, but modelled peaks • Na vs Cl measured: atmospheric origin. • BC concentration: • Overestimation rarely observed with ForSAFE. • BC deposition: the highest for EPC08. A Na (µeq.l-1.month-1) Years B BC (µeq.l-1.month-1) Na (µeq.l-1.month-1) Years Soil solution (-20cm) comparison between ForSafe monthlymodelled data and measured Renecofor data from 1993 to 2009 for Na (A) and BC (B). Cl (µeq.l-1.month-1)

  31. ForSAFE inputs A • Meteo France and RENECOFOR comparison of climate data. • Meteo France spatialised data SAFRAN, based on observed data and meteorological modelisation. • RENECOFOR : site Meteo station. Taverage (°C) Years B Tmax (°C) Years Monthly average temperature (A) and monthly max average T (B) (Meteo France and RENECOFOR CHS 41) from 1995 to 2008 (°C)

  32. ForSAFE evaluation Once calibrated, the model Veg can be coupled to ForSAFE outputs…

  33. Critical loads evaluation • Vegetation population modelled used to trace changes caused by Ndep. • Three variables have to be defined : • A reference population/time frame : • Population with vegetation under Background deposition • A target population chosen to be protected : • Total ground vegetation : 100 % • A limit for acceptable/excessive change : a critical limit! • 5 points of changes in the vegetation cover.

  34. Critical loads evaluation • How different can a population become, and in what time frame? • The critical limit: decision rule • 5 points of difference between the background vegetation composition and the vegetation considered. (m2cover/m2ground area) No vegetation species overlap, just replacement Veg composition and covering Belyazid et al., submitted

  35. Vegetation difference (points) Vegetation difference (points) Critical loads evaluation • The critical limit: decision rule • No time dependance • Average yearly exceedance • = area/years. Climit Climit (Belyazid et al., submitted)

  36. Average yearly difference (M) 5 Points BGD MFR CLE NOC CLN N Dep (mass/area,yr) Critical loads evaluation Critical load example Belyazid et al., submitted

  37. Critical loads evaluation kg.ha-1.yr-1 N Deposition NOC : highly exceeded CLE: CL still exceeded MFR : CL not exceeded MFR close to BGD Not much difference between sites except EPC08 Percentage of exceedance # for all sites N Critical load for a 5 points of vegetation composition changes for the 4 tested sites and with the 4 different deposition scenarii. Vegetation change : to be validated

  38. Some Messages/Questions…  N deposition (CLE) still have a potential impact on ecosystems  Coupling ecological response with soil response to N deposition : a powerfull approach but site calibration is needed for different situations  Soil response: vegetation indicators are neededed  Large Veg database : helpfull for coupled model in large areas but ALSO to take into account climate change

  39. Some Messages/Questions… • Model calibration needed : • -EMEP deposition • -VSD+ : local scale? country scale? • -N pool and C/N : evolution linked to carbon cycle? • -ForSafe-Veg: BC outputs, hydrology? • Evaluation of model sensibility, robustness •  Recovery : variable, effective for some sites • Historical landuse influence on forest biodiversity : • key point to evaluate future changes ?

  40. Role of historical land use on present-day forest biodiversity Dupouey et al., 2002 Where does biodiversity come from? Where does biodiversity go to?

  41. Acknowledgements • With the GREAT ! help of : • Jean-Claude Gégout, UMR AgroParisTech – ENGREF – INRA • Jean-Paul Party, Sol-Conseil Strasbourg • Jean-Luc Dupouey, INRA Nancy • Thierry Gauquelin, IMEP, Université de Provence • Didier Alard and Emmanuel Corcket, UMR BioGeco INRA Bordeaux • Vincent Boulanger, INRA Orléans and Jean-François Picard • Manuel Nicolas and Marc Lanier, RENECOFOR, ONF • Sophie Leguédois, CNRS, EcoLab, Toulouse • Harald Sverdrup and Bengt Nihlgård, Univ. Lund • Jaap Slootweg, Gert Jan Reinds, Luc Bonten • Laurence Galsomiès, ADEME Veg_table set up Patability RENECOFOR data Project set up Veg_parametrisation VSD+ helpers Financial support

  42. VSD+ input limitations • Shift between EMEP deposition and sites measurements : • Observed for each studied site • EMEP data used, • Overestimation of deposition. Eq.ha-1.yr-1 Years EMEP and RENECOFOR deposition comparison of SP57 RENECOFOR site for Ammonium (A), Nitrate (N).

  43. Forsafe inputs A B Dep NO3 (meq.m-2.y-1) Dep NH4 (meq.m-2.y-1) Years Years C Comparison of Annual deposition of NH4 (A), NO3 (B) and SO4 (C) from 1993 to 2008 (meq.m-2.y-1) between Renecofor (green) and EMEP scaled down (red). Dep SO4 (meq.m-2.y-1) Years

More Related