garber

Document Sample
garber Powered By Docstoc
					THE INFLUENCE OF
      STAND
 CONDITIONS ON
   TREE FORM
Sean M. Garber and Aaron R. Weiskittel
       Oregon State University
            June 21, 2004
         INTRODUCTION
• A goal of silvicultural research is to
  detect changes in growth.
• Changes in stand conditions affect
  diameter and sometimes height:
  – Density (spacing and thinning)
  – Social position within the stand
  – Control of competing vegetation
  – Fungal diseases
      INTRODUCTION

• Consequently stem volume
  also responds to these stand
  conditions.
        INTRODUCTION
• Existing volume and taper equations:
  – Most are functions of only DBH and HT;
  – Regional development;
  – May not include intensively managed
    stands or other dramatically altered stand
    conditions;
  – Averages lose subtle differences.
        INTRODUCTION
• Questions:
  – Is there an affect of stand conditions
    above and beyond the effect on DBH and
    height?
  – Does this influence the ability to discern
    growth differences among treatments in
    silvicultural studies?
      INTRODUCTION
HT




DBH
            METHODS
• Analytically assessed stem form on
  four separate studies:
  – Two vegetation management studies
  – A western larch LOGS thinning study
  – Two mixed-species spacing trials
  – Study of Swiss needle cast
             METHODS
• Kozak variable-exponent taper model
  applied in all studies:
                  Represents shape



    d = γ1   DBH γ2     XC    +ε
           METHODS
• Shape component (XC)
 –X varies from 0 (top), 1 (p), > 1
  below p.
 – C = linear function of:
   • Z = stem position (h/HT)
   • DBH/HT
   • Treatment variables
                Model dynamics
    1.0
    0.8




                                          Positive term
    0.6
Z

    0.4




                Negative term
    0.2
    0.0




          0.0   0.2   0.4   0.6     0.8   1.0   1.2
                            d/DBH
           METHODS

• Sample trees were climbed or
  felled:
  – Sampled across range of within
    treatment DBH;
  – DBH, HT, HCB, and CW measured;
  – Upper stem diameters (ib and ob)
    and height measured.
             METHODS
• Autocorrelation accounted for in two
  ways:
  – Model fit by tree and parameters modeled
    using SUR.
  – Model fit using GNLS with CAR(1)
• Multicollinearity was also a problem
  – Screened to reduce VIF
             Douglas-Fir Stem Taper
             Under Early Vegetation
                    Control


Study part of the Vegetation Management Research Cooperative, Oregon State University.
       INTRODUCTION

• As trees develop under intensive
  management:
  – Form differs from a cone (/12)D2H,
  – Trees large enough for existing
    volume or taper equations.
Summit

         Marcola
 Vegetation Management sites
• Study description
  –   Planted in 1993,
  –   Completely randomized design,
  –   8 treatments,
  –   3 replicates,
  –   Plot area = 0.112 ac,
  –   49 seedlings planted at 9.8 ft square spacing.
                 METHODS
• Treatments
  –   No herbicide,
  –   4 ft2 full control,
  –   16 ft2 full control,
  –   36 ft2 full control,
  –   64 ft2 full control,
  –   100 ft2 full control,
  –   100 ft2 woody vegetation control,
  –   100 ft2 herbaceous vegetation control.
              METHODS
• Diameter outside bark was collected at tree
  base, breast height, 8 ft, and every 4 ft
  above 8 ft.
• Observed DOB’s were modeled using
  variable-exponent model using GNLS.
• C was a function of site, area treated, and
  target vegetation.
                RESULTS
• Modeling
  – Model fits were excellent (R2 > 0.95);
  – Impact of autocorrelation eliminated w/
    CAR(1);
  – Multicollinearity present, albeit small.
• Treatment variables
  – Area treated and herb control significant
                                         Check                                                Check
      1 .0


                                         2-ft                                                 Herb
                                         5-ft                                                 Woody
      0 .8
      0 .6
h/H
      0 .4
      0 .2
      0 .0




             0 .0   0 .2   0 .4   0 .6   0 .8    1 .0   1 .2      0 .0   0 .2   0 .4   0 .6   0 .8   1 .0   1 .2

                                                               d/DBH
           Stem Taper in a Western
          Larch Levels-of-Growing-
            Stock Thinning Study


From Lennette A.P. 2000. Twenty-five-year responses of Larix occidentalis stem form to five stand
density regimes in the Blue Mountains of eastern Oregon. M.S. Thesis, Oregon State University. 59 p.
             METHODS
• Study description
  – Established in 1933,
  – Completely randomized design,
  – 5 treatments,
  – 2 replicates,
  – Plot area = 0.4 ac,
  – Thinned in 1966, 1975, and 1985
                 METHODS
• Treatments based on bole surface area
  (BSA) targets:

  BSA (ft2/ac)      SDI       Relative density
     5,000           65            0.16
    10,000          109            0.27
    15,000          150            0.37
    20,000          193            0.47
    25,000          235            0.57
          METHODS

• Data analysis:
  –Parameters fit by tree;
  –Parameters tested among
   treatments by MANOVA;
  –Parameters modeled using SUR.
              RESULTS
• Stem form responded to thinning:
  – C = a1sin-1(Z) + a2Z2;
  – MANOVA suggested differences among
    the treatments;
  – Parameters increased w/ thinning
    intensity:
    • a1 = f (DBH/HT)
    • a2 = f (CR)
                BSA: 15,000 to 25,000




BSA 10,000



    BSA 5,000
            RESULTS

• CR improved model fit after
  accounting for DBH/HT

• Treatment variables did not
  account for additional variation
  beyond DBH/HT and CR.
Increasing thinning intensity
        Stem Taper in Two Mixed-
          Species Spacing Trials


From Garber, S.M. and Maguire, D.A. 2003. Modeling stem taper of three central Oregon species
using nonlinear mixed-effects models and autoregressive error structures. For. Ecol. Manage. 179:
507-522.
                 Study sites
• Two study sites located at Pringle Falls EF
  on the Deschutes National Forest:
  – Pringle Butte (34 years)
     • Lodgepole pine and Ponderosa pine
     • 4500 feet elevation, west aspect, and dry
     • SI100 = 100 ft (Barrett 1978)
  – Lookout Mountain (27 years)
     • Grand Fir and Ponderosa pine
     • 5100 feet elevation, east aspect, and “more mesic”
     • SI100 = 110 ft (Barrett 1978)
         Experimental Design
• Completely randomized split-plot design
                                                         MX
   – Whole plot factor: spacing
                                                   P      F    P
   – Split-plot factor: species composition
                                                   F      P    F
                                                   P      F    P
• Replacement series setup
   – Only one mixture 50:50
                                              MX
                          Whole plot          GF
                                              PP       Split plot
                         INTRODUCTION

       • Mixed-species stand development:
             – Spacing
                  • More spacing, less stratification
             – Species composition
                  • PP over GF
                  • LP over PP
From Garber, S.M. and Maguire, D.A. 2004. Stand productivity and development in two mixed-
species spacing trials in the central Oregon Cascades. For. Sci. 50: 92-105.
                            Top height growth across spacing
                 12   (a)                  1 .8 - m   (b)                    3 .7 - m   (c )                  5 .5 - m
                                 LP                                LP                               LP
                                 PP                                PP                               PP
                 10
                 8
                 6
                 4
Top height (m)
                 2
                 0




                      (d)                  1 .8 - m   (e)                    3 .7 - m   ( f)                  5 .5 - m
                 12




                                 GF                                GF                               GF
                                 PP                                PP                               PP
                 10
                 8
                 6
                 4
                 2
                 0




                      0     10        20   30         0       10        20   30         0      10        20   30

                                                            Stand age (yr)
               RESULTS
• Modeling
  – Model fits were excellent (R2 > 0.95);
  – Impact of autocorrelation eliminated w/ random
    effects and CAR(1).
• Profiles
  – Showed differences among spacing;
  – Slight differences between pure and mixed
    stands.
                                                    SPACING
              Pin u s c o n to r ta                                        Ab ie s g r a n d is

                                                           1.8                                                         1.8
      1 .0




                                                           3.7                                                         3.7
                                                           5.5                                                         5.5
      0 .8
      0 .6
h/H
      0 .4
      0 .2
      0 .0




             0 .0       0 .2          0 .4   0 .6   0 .8   1 .0   1 .2   0 .0       0 .2          0 .4   0 .6   0 .8   1 .0   1 .2

                                                                  d/DBH
                             SPECIES COMPOSITION
                    1.8-m spacing                                         5.5-m spacing
                                                   Pure PP                                               Pure PP
      1 .0


                                                   Mix PP                                                Mix PP
                                                   Pure LP                                               Pure LP
                                                   Mix LP                                                Mix LP


                               Mix PP (subordinate)
      0 .8
      0 .6
h/H




                    Pure PP
      0 .4
      0 .2
      0 .0




             0 .0     0 .2   0 .4   0 .6   0 .8   1 .0   1 .2      0 .0     0 .2   0 .4   0 .6   0 .8   1 .0   1 .2

                                                                d/DBH
                       SPECIES COMPOSITION
                    1.8-m spacing                                         5.5-m spacing
                                                   Pure PP                                               Pure PP
      1 .0



                                                   Mix PP                                                Mix PP
                                                   Pure GF                                               Pure GF
                                                   Mix GF                                                Mix GF


                                           Pure PP
      0 .8
      0 .6
h/H
      0 .4




                      Mix PP
      0 .2




                    (dominant)
      0 .0




             0 .0     0 .2   0 .4   0 .6   0 .8   1 .0   1 .2      0 .0     0 .2   0 .4   0 .6   0 .8   1 .0   1 .2

                                                                d/DBH
                        DOUGLAS-FIR
                      STEM TAPER & SNC



Part of the Swiss Needle Cast Cooperative, Oregon State University.

From Weiskittel, A.R. 2003. Alterations in Douglas-fir crown structure, morphology, and dynamics
imposed by Swiss needle cast in the Oregon Coast Range. M.S. Thesis, Oregon State University. 389
p.
             INTRODUCTION
• SNC disrupts normal needle physiology, leading
  to premature loss of foliage

• 10 yrs of extensive defoliation has dramatically
  altered Douglas-fir crown structure & morphology
   – modified age class structure and vertical location of
     foliage
   – reduced branching
   – increased crown recession rates
            HYPOTHESES
• loss of foliage has decreased stem
  increment in the crown

• increased crown recession rates modified
  stem form

• stem taper equations require additional SNC
  covariates
                 METHODS
• 105 trees from 31 plots destructively
  sampled
  –   DBH; 12 – 65 cm
  –   Relative density; 21.5 - 73.6
  –   Site index; 27.7 – 47.4 m @ 50-yr
  –   Foliage retention; 1.2 – 4.4 yrs

• 3 to 4 trees/plot & 10 to 15 discs/tree
                ANALYSIS
• utilized modified Kozak’s equation
  – includes Z, DBHZ, DBH/HT, FOLRET
  – little correlation between covariates
  – continuous AR1 error structure

• tree volume predicted using the developed
  equation and Bruce & Demars (1974)
  equation
  – results compared graphically and statistically
                  RESULTS
• R2 of 0.95

• parameters all highly significant (p<0.001)

• likelihood ratio test also suggested that foliage
  retention significantly improved model fit
  (p<0.0001)

• for a given DBH/HT, foliage retention
  significantly reduced dibs throughout the stem,
  except below BH
                     RESULTS
      1.0



      0.8                         HIGH SNC
                                  LOW SNC


      0.6
h/H




      0.4



      0.2



      0.0




            0   10        20         30      40

                       DIB (cm)
            DISCUSSION
• Does relative stem form change w/
  stand conditions?
  – Differences in profiles after accounting
    for DBH and HT.
  – Treatments variables significant after
    accounting for DBH/HT:
    • Vegetation treatment;
    • Thinning level;
    • Foliage retention.
               DISCUSSION
• Does relative stem form change w/ stand conditions?

   Study         Shape change        Consequences
↑ Veg. control More Neoloid         Overestimate
↑ Thinning     More Neoloid         Overestimate
↑ Spacing      More Neoloid         Overestimate
↑ Dominance    More Neoloid         Overestimate
↑ SNC          More Neoloid         Overestimate
            DISCUSSION
• Do these changes affect results of
  silvicultural studies?
  – Significant difference between volumes
    estimated from SNC equations and Bruce-
    DeMars;
  – Bias a function of top height, DF QMD,
    FOLRET, and CLSA;
  – Mean volume losses due to SNC were 31%.
            DISCUSSION
• Mechanisms
  – Differential basal area growth rate along
    the stem differences among treatments.
    • Stem growth rate responds to increased
      resources;
    • Response increases w/ decreasing stem
      height.
  – Swiss needle cast
    • Reduces tree leaf area (thins crown)
    • Decreases upper stem growth
Curves diverge
           DISCUSSION
• Measure upper stem diameters.
  – Assess existing volume or taper
    equations.
  – Develop new site-specific equations.
• Incorporate crown ratio into the model
• Include stand or treatment variables
  into models.
  THE END


Any questions?

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:1
posted:2/21/2012
language:
pages:51