Using the 3-PG Model to Predict Hybrid Poplar Yields in Minnesota

Report
Using Process-Based Modeling to
Inform SRWC Species Selection
and Management in the
Southeastern USA
William Headlee1
Richard Hall1
Ronald Zalesny, Jr.2
Matthew Langholtz3
1Iowa
State University Department of Natural Resource Ecology & Management, Ames, IA
2USFS
3DOE
Northern Research Station Institute for Applied Ecosystem Studies, Rhinelander, WI
Oak Ridge National Laboratory, Oak Ridge, TN
Modeling Short-Rotation Woody Crops
• Project overview
▫ Goals and objectives
▫ Model description
• Current progress
▫
▫
▫
▫
Literature review and data compilation
Model fitting
Mapping productivity
Economic analyses
Goals & Objectives
• Southeastern USA has suitable climate for a variety of SRWCs
▫ Which woody crops are best-suited for different combinations of
climate and soil conditions?
• Process-based models allow for comparisons of crops, in lieu
of extensive side-by-side testing
▫ 3-PG model already calibrated for three candidate SRWCs (poplars,
loblolly pine, and eucalypts)
Poplars
(photo R. Zalesny)
Loblolly pine
(photo W. Ciesla)
Eucalypts
(photo D. Haugen)
Goals & Objectives (continued)
• Primary objective:
Within range of overlap,
model productivity and
compare economics for
each species on
marginal lands.
• Additional objective:
Develop methods to
account for differences
in silvicultural practices
(site prep, weed control,
fertilization, etc).
Geographic ranges for (A) poplars, (B) loblolly pine, and (C)
eucalypts. Adapted from USDA Plant Hardiness Zone map.
Overview of 3-PG Model
• “Physiological Principles Predicting
Growth” model (Landsberg & Waring
1997) accounts for key drivers of growth
Sunlight
▫ Available resources (sunlight, water, nutrients)
▫ Tree physiology (resources → biomass)
• Uses site-specific inputs for climate and
soils to estimate available pools of key
resources
▫ Sunlight (solar radiation)
▫ Water (precipitation, temperature, soil water
holding capacity, water table depth, etc.)
▫ Nutrients (site fertility)
Water
Nutrients
Overview of 3-PG (continued)
• Species-specific physiological parameters
determine the amount of photosynthate
produced from available resources, and its
allocation to tree components
CO2
▫ Quantum canopy efficiency
▫ Respiration
Leaf
fall
▫ Leaf litterfall rate
▫ Root turnover rate
▫ Biomass partitioning (stem, branches, foliage,
bark, roots)
▫ The list goes on… 60 parameters in all!
Turn
over
Literature Review & Data Compilation
• Identified 57 previouslypublished growth and
yield studies for crops of
interest in MS, AL, FL, GA,
and SC
• Contain 441 unique
combinations of soils,
management, planting
density, etc. (stands)
Model Fitting
• Use previously-published data to validate
existing base calibrations for poplars
(Headlee et al 2013), loblolly pine (Bryars
et al 2013), and eucalypts (Sands 2004)
Sunlight
• Accounting for differences in silvicultural
practices
▫ Planting density – direct input
▫ Irrigation – roll into precipitation
▫ All others – quantify responses with
linear regression, roll into fertility rating
Water
Nutrients
Model Fitting (continued)
• Silviculture regression
model asssumptions
20% change
ininintercept
slope
and slope
▫ Increasing application
rates associated with
diminishing rates of
return
Cumulative Growth
▫ Practices can influence
slope and/or intercept of
cumulative growth curve
Time
Model Fitting (continued)
• Regression models based on typical cumulative growth function (Avery and
Burkhart 2002) where “G” is cumulative growth, “a” is the intercept (growth
ceiling), and “b” is the slope (sigmoid curve):
=
−
1

• To predict impacts of treated stands (subscript T) relative to untreated controls
(subscript C):
 −
1
1
    
( −  ) − ( −  )
∆a −∆b

=
=
=
1

 −
   
1

• Log-transform and separate ∆a and ∆b into components for fertilizer (F), weed
control (W), site preparation (S), pest control (P), and residuals (R):
ln

1
1
= ∆a − ∆b
= (∆aF + ∆aW + ∆aS + ∆aP + ∆aR ) – (∆bF + ∆bW + ∆bS + ∆bP + ∆bR )



Model Fitting (continued)
• Silviculture linear regression – results
▫
▫
Responses varied by SRWC and growth parameter
Regression model fit (r2) ranged from 0.57 to 0.88
Height Regression
∆ Intercept ∆ Slope
ns
Eucalypts
N
<0.0001
ns
P
<0.0001
Poplars
N
<0.0001
0.0091
ns
P
ns
ns
K
ns
Site Prep.
ns
<0.0001
ns
Loblolly
N
ns
P
<0.0001
ns
ns
K
ns
ns
Site Prep.
ns
Weed Ctrl.
<0.0001
0.0037
ns
Pest Ctrl.
ns
Crop
Practice
DBH Regression
∆ Intercept ∆ Slope
<0.0001
ns
ns
ns
<0.0001
ns
ns
ns
ns
ns
ns
<0.0001
ns
ns
ns
<0.0001
ns
ns
ns
ns
ns
<0.0001
ns
ns
Model Fitting (continued)
• Silviculture linear regression – examples
30
8
6
4
Predicted - Fertilized
Observed - Fertilized
Observed - Control
2
0
2.0
2.5
3.0
3.5
Age (years)
4.0
Poplar DBH (cm)
Eucalypt DBH (cm)
10
20
10
Predicted - Site Prep
Observed - Site Prep
Observed - Control
0
0
5
10
Age (years)
15
Loblolly Pine DBH (cm)
▫ Eucalypt response to fertilization (Rockwood et al
2008)
▫ Poplar response to site preparation (Francis 1982,
Baker and Blackmon 1978)
▫ Loblolly pine response to fertilization (Samuelson
et al 2008 & 2004)
30
20
10
Predicted - Fertilized
Observed - Fertilized
Observed - Control
0
0
5
10
Age (years)
15
Model Fitting (continued)
• Next step: validate previous model calibrations using
growth data compiled from the literature
▫ Fertility ratings estimated from regression models
▫ Climate data (temperature, precipitation, etc.) from
NOAA National Climatic Data Center
▫ Soils data (water holding capacity, depth to water table,
etc.) from NRCS soils database (SSURGO)
Mapping Productivity
• Similar to previous poplar modeling,
will focus efforts on suitable lands
• “Marginal lands” as defined by NRCS
Land Capability Classes (LCC) II-IV
▫ Moorhead & Dangerfield (1998) found
>95% of ag lands converted to woody
crops in Georgia were LCC II-IV
Source: Zalesny et al 2012
LCC
I
II-IV
V-VIII
Definitions (USDA-NRCS 2013)
Soils have slight limitations that restrict their use
Moderate to very severe limitations restrict crop choices, require special conservation
practices, or both
Unsuited to cultivation; limitations preclude commercial plant production
Economic Analyses
• Model’s productivity estimates will be used to
compare:
▫ Mean annual increments (MAI) for each SRWC under
different combinations of climate and soils
▫ Optimum rotation ages (maximum MAI)
▫ Land expectation values (LEVs)
▫ Most economically feasible SRWC (maximum LEV)
Summary
• Literature Review
▫
Identified 57 studies containing sufficient
data to model 441 unique combinations of
site conditions and silvicultural practices
• Model Fitting
▫
▫
Used linear regression to quantify SRWC
responses to silvicultural practices
Validating model with published yield data
using fertility rating estimates, NOAA
climate data, and NRCS soils data
• Mapping Productivity
▫
Will focus on marginal lands (NRCS Land
Capability Classes II-IV)
• Economic Analyses
▫
Will compare crop MAIs, optimum rotation
ages, and LEVs
References
Avery TA, Burkhart HE (2002) Growth and yield models. In: Forest Measurements, 5th ed. McGraw-Hill, New
York, pp 352-385
Baker JB, Blackmon BH (1978) Summer fallowing – a simple technique for improving old-field sites for
cottonwood. Forest Service Research Paper SO-142. 5 p.
Bryars C, Maier C, Zhao D, Kane M, Borders B, Will R, Teskey R (2013) Fixed physiological parameters in the 3-PG
model produced accurate estimates of loblolly pine growth on sites in different geographic regions. Forest
Ecology and Management 289: 501-514
Francis JK (1982) Fallowing for cottonwood plantations: benefits carry to rotation’s end. In: Proc. North
American Poplar Council 19th Annual Mtg. p 1-7.
Headlee WL, Zalesny RS, Donner DM, Hall RB (2013) Using a process-based model (3-PG) to predict and map
hybrid poplar biomass productivity in Minnesota and Wisconsin, USA. BioEnergy Research 6: 196-210
Landsberg JJ, Waring RH (1997) A generalised model of forest productivity using simplified concepts of
radiation-use efficiency, carbon balance and partitioning. Forest Ecology and Management 95: 209-228
Rockwood DL, Carter DR, Stricker JA (2008) Commercial tree crops for phosphate mined lands. University of
Florida Publication 03-141-225. 86 p.
Samuelson LJ, Butnor J, Maier C, Stokes TA, Johnsen K, Kane M (2008) Growth and physiology of loblolly pine in
response to long-term resource management: defining growth potential in the southern United States.
Canadian Journal of Forest Research 38: 721-732.
Samuelson LJ, Johnsen K, Stokes T (2004) Production, allocation, and stemwood growth efficiency of Pinus
taeda stands in response to 6 years of intensive management. Forest Ecology and Management 192: 59-70.
Sands PJ (2004) Adaptation of 3 PG to novel species: guidelines for data collection and parameter assignment.
Technical Report 141, CRC for Sustainable Production Forestry, Hobart, Australia.
Zalesny RS, Donner DM, Coyle DR, Headlee WL (2012) An approach for siting poplar energy production systems
to increase productivity and associated ecosystem services. Forest Ecology and Management 284: 45-58
Thank you for your time!
• Funding and other support provided by:
▫ USDA-NIFA Agriculture and Food Research Initiative (CRIS 2013-03276)
▫ USFS-NRS Institute for Applied Ecosystem Studies (IAES)
• Technical support
▫ Sue Lietz (IAES)
• Literature review assistance/suggestions
▫ Dave Coyle
▫ John Stanturf
▫ Lynne Wright
• Questions?

similar documents