Interactions between C and N cycles in terrestrial ecosystems: progress and uncertainties

Transcription

1 Interactions between C and N cycles in terrestrial ecosystems: progress and uncertainties M. Mencuccini School of GeoSciences University of Edinburgh (UK)

2 OUTLINE Brief review of N cycle and links to C cycle Alteration of N cycle by N deposition Impacts of N deposition on C cycle Progress and uncertainties

3 A BRIEF REVIEW ON N CYCLE Leaching Denitrifying: NO 3 -NO 2 -NON 2 ON 2 Probably most important nutrient (with P) Ultimately derived from the atmosphere by lightning strikes or biological N fixation N fixation by free- living or symbiotic bacteria Fundamental role of microbes in N cycle

4 Forests are very good at keeping hold of nitrogen

5 Streamwater nitrate concentrations (deforested vs control catchment)

6 N-limitation are common in Boreal forests (Hogberg 2007) courtesy Sune Linder, Umea Univ

7 ##!"

8 Interactions with elevated CO 2 Young stands [CO 2 ] = 550 ppm 2050 NPP Soil metabolism Norby et al, 2005

9 Ndep level at Oak Ridge (TN) 15 kg N ha-1 yr-1

10 Alterations in N cycle

11 Nitrogen sources REDUCED N (NH 3 ) OXIDISED N (NO, NO 2, HNO 3 ) 25-30% 25-30% Electricity generation Globally 50-70% of fixed N is anthropogenic Traffic Industry

12 No/little progress in abatement

13 N deposition, yrs 2000 and 2030 N input to ecosystems doubled since the beginning of the industrial revolution, uneven distribution Forecasts: small reduction in Europe and N. America, strong increase in developing countries mmol N m -2 yr -1 (Galloway et al. 2004)

14 Complete accounting for GH effects + Atmospheric ozone + N leaching, water quality Biodiversity acidification N use and emissions N deposition on natural ecosystems Aerosol, CH 4 turnover + N 2 O Ecosystem C sink + _? Climate change Other environmental effects Climate forcing

15 Radiative forcing of N deposition Long-lived greenhouse gases Fossil fuel & landuse change CO [382 to 469] -74 [ ] biospheric CO 2-19 [ ] (incl. atmos. fertilisation & O 3 effect) 4.4 [ ] CH 4 (decreased atmospheric lifetime & 24.5 [22-27] and decreased soil uptake) -4.6 ( ) 0.13 [ ] N 17.0 [ ] 2 O 17.0 [ ] Halocarbons 7.5 [ ] Ozone Stratospheric water vapour from CH 4 Surface albedo Stratospheric Tropospheric Land use Black carbon on snow 8 < [ ] 2.9 [ ] 3.6 [1.0 to 6.1] [ ] 9.9 [0-19.8] Total aerosol Direct effect Cloud albedo effect Linear contrails Total Anthropogenic Sulphates [-16.5 to -36.5] -5.4 [-9.4 to -1.4] (SO 2 oxidation & aerosol neutralisation) Nitrate [ ] [ ]?? < [ ] [ ] European contribution to global radiative forcing [mw m -2 ] dc/dn= 0 vs 200 kg C / kg N

16 Impacts of N deposition on C Cycle: concerns over forest health in the 1980s Reductions in sulphur emissions Other reasons?

17 Aber et al., 2003

18 Aber s hypothesis of N saturation Aber et al (1998) Bioscience Saturation around here

19 Pollution studies Forest inventories Eddy covariance

20 Existing estimates of dc/dn METHOD Tree growth in 7 Ndep experiments 15 N tracer and C/N ratios (tree and soil) Upscaled tree growth and soil (121 plots) Fertilisation studies (17 studies) dc : dn (kg/kg) ~ 0 : 1 ~ 46 : 1 ~ 48 : 1 ~ :1 AUTHORS Emmett (1999) Nadelhoffer et al. (1999) De Vries et al. (2006) Hyvonen et al. (2007) Total tree growth (20,067 plots) ~ 73 : 1 Thomas et al (2010) NEP from eddy covariance or chronosequences (54 studies) ~ 200 : 1 Magnani et al (2007)

21 How to resolve uncertainties? Theory and modelling Independent observations Old and new experiments

22 Current models Predict significant positive role of N deposition on C cycle Tend to agree with relatively low estimates of dc/dn Do not necessarily contain all important processes

23 New version of G Day Increase in Ndep + management N retention by canopies included (50% of Ndep) Management included (2.5% yr -1 ) Change in root allocation included Dezi et al (2010) GCB

24 Long-term changes in soil C:N N deposition Soil C:N Net N mineralisation Decreasing soil C/N Wood production, NEP Time

25 Limitations in modelling Results dependent on which processes are represented How do we select the important processes? How do we know that including other processes would not reverse the trends? Hypothesis driven modelling: making the point.

26 Independent observations Over space Over time Different scales

27 Limitations in observations How well do we know N deposition? How well do we know NEP? Single vs multiple regression problem (or statistical vs process-based modelling) Confounding factors It s s the meta-analysis, analysis, stupid!

28 N addition experiments (e.g., Hyvönen et al. 2007) 50 (trees) + 20 (soil) kg C / kg N (e.g., Emmett 1999)

29 Limitations in (N addition) experiments Simulated Ndep by-passes the canopies Doses are too high Times are too short Must account for existing Ndep (is dc/dn a useful index?)

30 METHOD Tree growth in 7 Ndep experiments 15 N tracer and C/N ratios (tree and soil) Upscaled tree growth and soil (121 plots) Fertilisation studies (17 studies) Total tree growth (20,067 plots) NEP from eddy covariance or chronosequences (54 studies) What is dc/dn? dc : dn (kg/kg) ~ 0 : 1 ~ 46 : 1 ~ 48 : 1 ~ :1 ~ 73 : 1 ~ 200 : 1 AUTHORS Emmett (1999) Nadelhoffer et al. (1999) De Vries et al. (2006) Hyvonen et al. (2007) Non sensical? Thomas et al (2010) Magnani et al (2007)

31 OVERALL CONCLUSIONS There is little doubt that N deposition is affecting the C cycle and increase sequestration Magnitude of this effect is uncertain dc/dn probably not a reasonable index of sensitivity of carbon cycle to Ndep How to approach uncertainty: more data required but, especially, more ideas! Acknowledgements F. Magnani, M. Borghetti, D. Stevenson, T. van Noije, S. Raddi, J. Grace