Modelling Mixedwoods at the Whole-Stand Level

1
Modelling Mixedwoods at the Whole-Stand Level

• Oscar García
• University of Northern British Columbia

2
Outline

• Why??
• Spatial structure
• Initial state info
• Predictability
• How?
• Canopy driven
• Species layers, interaction
• Model structure

3
Why??!
4
Spatial structure

• Tree sizes not random on the ground
• Competition ? neighbours more different
• Micro-site ? neighbours more similar
• ? Size distribution properties change with area

5
Spatial structure
Expected dbh variance in a circular plot
6
Initial state estimates
Samples of 50
7
Limits to predictability
8
Limits to predictability
9
Limits to predictability
10
Limits to predictability
11
Limits to predictability
12
Limits to predictability
13
Limits to predictability
14
Aggregation
PV kT
d2x / dt2 F/m
(Newton, 1687)
15
Understanding, prediction
Explain
Predict?
16
Simulation
Prediction?
17
Predictability
18
Complex systems requiresimple models
19
Mixedwoods
20
Individual-tree?
Tree-level model
tree list
tree list
21
Individual-tree?
Tree-level model
tree list
tree list
(B,N,H)
Inventory
22
Individual-tree?
Tree-level model
tree list
tree list
(B,N,H)
(B,N,H)
Inventory
Application
23
Individual-tree?
Tree-level model
tree list
tree list
Stand-level model
(B,N,H)
(B,N,H)
Inventory
Application
24
Whole-stand

• Mix species, uneven-aged Eg. Moser
• What does the xylem have to do with it?
• Allometry, or lack of it
• Top down

25
Aspen-spruce
26
Aggregated (whole-stand)
Aspen foliage
Aspen wood
Spruce foliage
Spruce wood
27
Mechanism
Interceptance ()
Volume (m3/ha)
28
Site quality
29
Site quality
30
Gross increment
31
Resource capture

• Closure
• ? Amount of foliage, relative to maximum
• Occupancy (R)
• ? Interceptance, relative to maximum

R
Closure
R 1- (1- C)2.4
32
Simplest models

• First approximation
• dV/dt a (closed canopy, no mortality)
• Including open, no mortality
• dV/dt a R
• dR/dt b (1 R) or b R(1 R)
• Mortality
• dN/dt - c R N dV/dt a R - Vmort
• Vmort (mean V of dead) (- dN/dt)
• - k N-1/2 V/N dN/dt

33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
Conclusions

• Predicting behaviour of individual trees may be
  hopeless
• Not necessary
• No dbh-driven modelling
• More research is needed