Developing a Whole-stand Model for Douglas-fir Plantations

1
Developing a Whole-stand Model for Douglas-fir
Plantations

• Eric C. Turnblom
• Samuel D. Pittman

2
Whole-stand Model Topics

• Introduction
• Model Development
• Biological / Model Theory
• Equations / Data / Fitting
• Fit Assessment / Model evaluation
• Model Uses Future Work
• Conclusion

3
Introduction

• Stand Management Cooperative
• Mission
• To provide a continuing source of high-quality
  information on the long-term effects of
  silvicultural treatments and regimes on stand and
  tree growth and development and on wood and
  product quality

4
Introduction

• Stand Management Cooperative
• Objectives
• To support modeling
• Installations contain a wider range of treatments
  than current practice
• To support managers
• Provide timely information and guidance to those
  who are managing similar stands

5
Stand Management Cooperative

• Membership
• Companies Boise Cascade, Longview Fiber, Port
  Blakely Tree Farms, West Fork Timber, etc. (15 in
  total)
• Tribal Lands Quinault Dept. Natural Resources
• Federal Agencies Bureau of Land Management
• State County Agencies ODF, WA DNR, King Co.
  DNR
• Institutions BC Ministry Of Forests, FORINTEK
  Canada, USFS PNW Research Station
• Universities OSU, UBC, UW
• Suppliers Agrium UJS, Inc., JR Simplot Co., etc.

6
Stand Management Cooperative

• Organization
• Policy Committee (chair, vice-chair)
• Director
• Projects, Field Crew, Database Management, Office
  Staff
• Project (Technical Advisory Committees, or TACs)
• Nutrition
• Silviculture
• Wood Quality
• Modeling

7
Stand Management Cooperative

• Whats measured
• Tree and derived Stand Attributes
• Live Species, DBH, Height, height-to-crown base,
  crown width, damage
• Dead size condition of snags until fallen
• Branch Measurements (BH)
• Largest Branch diameter, branch count (live
  dead)
• Understory Vegetation
• Average foliage height, coverage () by species

8
Stand Management Cooperative

• Whats Measured
• Habitat Indicators
• Depth of duff to mineral soil, FSD ()
• Site Characterization
• Soil
• 0 - 6, 6 - 12 composited from five sample
  points per plot
• Forest Floor
• 1 sq. ft plots, four per measurement sample plot
• Foliar Nutrients
• Upper 1/3 of crown, 1 year after fertilization,
  12 trees per plot

9
Stand Management Cooperative

• History
• Regional Forest Research Nutrition Project
  (RFNRP)
• Commenced in 1969 (western WA, OR)
• Stand Management Cooperative (SMC)
• Formed 1985, 1st operational year 1986
• Western WA, OR, southwestern BC
• In 1991, RFNRP rolled into SMC Nutrition Project

10
Stand Management Cooperative

• Cohorts
• RFNRP
• Phase I Unthinned natural stands, originating
  1930s
• Phase II Thinned natural stands, originating
  1930s
• Phase III Young thinned plantations, planted
  1940s
• Phase IV Precommercially thinned plantations
  1960s
• SMC
• Type I Variously treated plantations 1970s
• Type II Variously thinned plantations 1950s
• Type III Range of planting spacings late 80s
  - 90s
• Type IV Range of planting spacings, genetics
  2Ks

11
Introduction

• Study Modeling Objectives
• Benchmark existing growth simulators
• Summarize early growth development of immature,
  managed stands of Douglas-fir
• Investigate and understand the influence of stand
  structure on growth and yield
• Produce variable-density growth and yield tables
  in user-friendly format

12
Whole-stand Model for Douglas-fir

• Model Development
• Biological / Model Theory
• Equations / Data / Fitting
• Fit Assessment / Model evaluation
• Model Uses Future Work
• Conclusion

13
Model Development

• Biological Theory
• Growing space
• As available space decreases, so does potential
  for growth
• Allometric relationships
• Relate growth in part of an organism to another
  part
• Maximum (limiting) size - density relationships
• Only so much biomass can be packed into
  available space

14
Biological Theory

• Maximum (limiting) size density relationship

(Xue, et al. 1999)
15
Biological Theory

• Every stand has a unique self-thinning boundary
  line driven by its unique stand allometry

(Turnblom Burk 2000)
16
Biological Theory

• Stand allometry is dynamic, not static
  allometric relationships change with stage of
  development
• The rate at which a stand variable changes is
  found by differencing gross growth respiration
  (anabolic catabolic processes, respectively)
• The magnitudes of the anabolic and catabolic
  processes depend on the current state of the
  stand and is affected by site quality

17
Model Theory

• The current state of the stand can be described
  by the triplet (stand basal area, quadratic mean
  DBH, and dominant height)
• Dynamic allometry is captured through
  simultaneous, coupled rate equations

18
Model Equations
19
Model Fitting Data

• Untreated, pure DF (gt 80 by BA) plots
• Initial BHA averaged 14 years, ranging to 51
  years
• Site index from 85 to 145 ft _at_ 50 yrs BHA (King)
• 459 plots total 249 used in fitting, 210 for
  benchmarking
• Measured every 4 yr, ranging 1 to 6, averaging
  4.4
• Single plot measurements span 4 to 26 yrs

20
Model Fitting Parameterization

• First order, centered differencing used to create
  a differential data set matching a yield data set
  constructed from mid-point of yield values
• System of three equations was fit to the data
  using the objective function

21
Fit Evaluation

• Residual graphs were examined
• Visual quality of predicted trajectories were
  assessed for agreement with known biological
  behaviors
• Attempts were made to augment the system of
  equations with other system state variables (e.g.
  stand form)

22
Fit Evaluation

• Residual plotting

23
Fit Evaluation

• Residual plotting

24
Model Evaluation

• Results - biological behavior
• Growth yield dynamics on two sites
• High site (dashed) SI 145
• Low site (solid) SI 85
• Otherwise similar in Q, B, H, Age

25
Model Evaluation

• Biological behavior

26
Model Evaluation

• Biological behavior

27
Model Evaluation

• Biological behavior

28
Model Evaluation

• Trajectory behavior

29
Whole-stand Model for Douglas-fir

• Model Uses Future Work
• Conclusion

30
Model Uses Future Work

• Boundary conditions to individual-based or
  tree-list models
• TreeLab (v1.0)
• Variable-density yield simulator for immature,
  managed Douglas-fir plantations
• Site productivity evaluation modeling in
  management-oriented and process-based growth
  model hybrids (proposal to USFS / AFPA collab.)

31
TreeLab (v1.0)
32
TreeLab (v1.0)
33
Site productivity modeling

• This proposed research seeks to
• Better understand mechanisms driving forest
  growth productivity in young DF stands
• Examine interpret association between
  environmental HydroGeoMorphic attributes of a
  site / stand and parameters of the TreeLab
  whole-stand model
• High potential to derive a site-characterization
  parameter to augment site index

34
Conclusion

• Whole-stand Model propositions
• Unique self-thinning boundary for each stand
• Stand allometry is dynamic, not static
• Changes in stand attributes are represented as a
  coupled system of rate equations that govern
  allometry
• Whole-stand model with the most plantation data
• Young data set (for stand development)
• Based on assumed maximums (BA, QMD and Htop)

35
Conclusion

• Please visit lthttp//www.standmgt.org/gt
• To download, click on the TREELAB link (LHS)
• Eric C. Turnblom ltect_at_u.washington.edugt