estimating litter and soil carbon stocks on managed forest land …€¦ · gelisol 0 10 63...
TRANSCRIPT
Estimating litter and soil carbon stocks on managed
forest land in Alaska, USAGrant M. Domke1, Andrew Gray2, Brian Walters1, Charles Perry1, Lucas Nave4, Stephen Ogle3, Brendt Mueller2
1Northern Research Station, USDA Forest Service; 2Pacific Northwest Research Station, USDA Forest Service; 3Natural Resource
Ecology Laboratory, Colorado State University; 4University of Michigan Biological Station
1
Importance of C in soils
Largest terrestrial sink
UNFCCC reporting
Much is in forests
Influences biological, chemical, and
physical soil functions
…much uncertainty exists
Jandl, R., Rodeghiero, M., Martinez, C., Cotrufo, M. F., Bampa, F., van Wesemael, B., Harrison, R.B., Guerrini, I.A., deB Richter Jr., D., Rustad, L., Lorenz, K., Chabbi, A., Miglietta, F. 2014. Current status, uncertainty and future needs in
soil organic carbon monitoring. Science of the Total Environment, 468, 376-383.
Lal, R. 2005. Forest soils and carbon sequestration. Forest Ecology and Management, 220(1), 242-258.2
Why the change in litter and SOC methods?
Smith, James E. et al. 2006. Methods for calculating forest ecosystem and harvested carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square, PA: U.S. Department of Agriculture,
Forest Service, Northeastern Research Station. 216 p
Amichev, B. Y., & Galbraith, J. M. 2004. A revised methodology for estimation of forest soil carbon from spatial soils and forest inventory data sets. Environmental Management, 33(1), S74-S86.
3
Domke, G.M., Walters, B.F., Perry, C.H., Woodall, C.W., Russell, M.B., Smith, J.E. 2016. Estimating litter carbon stocks on forest land in the United States. Science of the Total Environment 557-558: 469-478.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L., Swanston, C. 2017. Toward inventory-based estimates of soil organic carbon in forests of the United States Ecological Applications.
NFI sampling frame
Core field
1 plot per 2,430 ha
Intensive field
1 plot per 38,880 ha
Prefield analysis
Bechtold W.A. and Patterson, P.J. 2005. The enhanced Forest Inventory and Analysis program—national sampling design and estimation procedures. Gen. Tech. Rep. SRS-80. 85 p.
4
Managed forest land in Alaska
5
NFI plot design
Bechtold W.A. and Patterson, P.J. 2005. The enhanced Forest Inventory and Analysis program—national sampling design and estimation procedures. Gen. Tech. Rep. SRS-80. 85 p.
6
NFI litter and soil sampling
ucftBDCSOC iiiFIA
USDA Forest Service 2011. Phase 3 Field Guide – Soil Measurements and Sampling, Version 5.1. Last accessed: 08/10/2015. http://www.fia.fs.fed.us/library/field-guides-methods-proc/docs/2012/field_guide_p3_5-
1_sec22_10_2011.pdf
O’Neill, K.P., Amacher, M.C., Perry, C.H. 2005. Soils as an indicator of forest health: a guide to the collection, analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis program. Gen. Tech. Rep. NC-258. St.
Paul, MN: US Department of Agriculture, Forest Service, North Central Research Station. 53 p.
7
NFI sampling and reporting depth
Belowground C without coarse roots but
including fine roots and all other organic
carbon not included in other pools, to a
depth of 1 m.
Smith, James E.; Heath, Linda S.; Skog, Kenneth E.; Birdsey, Richard A. 2006. Methods for calculating forest ecosystem and harvested carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p
8
NFI and ISCN data
SOIL_ORDER FIA profiles ISCN profiles Measurements
Alfisol 840 558 5895
Andisol 135 154 1256
Aridisol 129 9 281
Entisol 234 127 1190
Gelisol 0 10 63
Histosol 58 171 1203
Inceptisol 577 250 2614
Mollisol 588 200 2451
ORDER_NOT_FOUND 0 265 2254
Oxisol 0 13 82
Spodosol 407 246 2583
Ultisol 686 311 3553
Vertisol 13 11 114
TOTAL 3667 2325 23539
9
NFI and ISCN data
10
lm fits
DepthISOC 1010 loglog
Soil order Intercept Slope r2 F-statistic p value
All 1.1795 -0.8228 0.56 29646.79 <0.001
Alfisols 1.1122 -0.8330 0.64 10657.50 <0.001
Andisols 1.3837 -0.8425 0.49 1185.78 <0.001
Aridisols 0.2065 -0.1300 0.02 6.55 0.011
Entisols 0.9300 -0.7207 0.39 752.34 <0.001
Histosols 1.6227 -1.0109 0.59 1724.22 <0.001
Inceptisols 1.1631 -0.7331 0.52 2833.00 <0.001
Mollisols 1.0163 -0.6214 0.51 2569.03 <0.001
Spodosols 1.4262 -0.9801 0.61 4097.61 <0.001
Ultisols 1.1576 -0.8867 0.68 7450.16 <0.001
Vertisols 0.5145 -0.2427 0.08 9.58 0.002
11
Domke, G.M., et al. 2017. Toward inventory-based estimates of soil organic carbon in forests of the United States. Ecological Applications
Predicted profiles
12
Domke, G.M., et al. 2017. Toward inventory-based estimates of soil organic carbon in forests of the United States. Ecological Applications
From 448 plots to 3816 plots…
13
Modeling approaches
LAT, LON, ELEV, SLOPE, ASPECT, WATERCD, FORTYPGRP, STDAGE, PHYSCLCD, SITECLCD, STDORGCD, DISTURB,
TRTMNT, BALIVE, SDI, REL_DENS, AGLT_C_MGHA, WORLDCLIM (many), GMI, YSD, SOIL_ORDER, SURF_NAME,
SURF_DEPTH, FIRE_PERIMETER, PCT_GELISOL, PERMAFROST_CLASS, NDVI (many), LANDFIRE (many), among
others
Breiman L. 2001. Random forests. Machine Learning. 45(1):5-32.
14
Pruning and variable selection
),,max,,,,,,()( surfgeoordergmitpptfortypgrpelevlonlatfSOCP
RMSE: 34.9 on 1091 DF,
R2: 0.38,
F-stat: 536 on 1 and 1091 DF,
p-value: < 0.0001
15
Domke, G.M., et al. 2017.. Toward inventory-based estimates of soil organic carbon in forests of the United States. Ecological Applications
Model comparisons for coastal AK
The estimates (NFI observations) and predictions (RF predictions) were statistically equivalent
Wellek, S. 2003. Testing Statistical Hypotheses of Equivalence. Chapman & Hall, London, England.
16
Soil carbon
Litter carbon
Take aways… The contribution of litter and soil C in forests has
been grossly underestimated in recent NIRs in the
US
Preliminary estimates of litter C stocks on
managed forest land in AK = 985 ± 153 MMT C –27% of all litter C on managed forest land in the
US
Preliminary estimates of soil C stocks (30 cm) on
managed forest land in AK = 3762 ± 513 MMT C –12% of all soil carbon stocks in the US
Improvement but much variability and substantial
uncertainty 17
Thanks!
Grant Domke: [email protected]
FIA carbon: http://www.fia.fs.fed.us/forestcarbon/
FIA program: www.fia.fs.fed.us
18