Limits and Possibilities for Sustainable Development in Northern Birch Forests:

AO Gautestad, FE Wielgolaski*, B Solberg**, I Mysterud*

* Department of Biology, University of Oslo, Norway

** Agricultural University of Norway, Department of Forestry, Ås, Norway

A general landscape-scale model relating mountain birch forest dynamics to various anthropogenic

influences, herbivory, and climate change

Outline

• Introduction: What can a model do for the HIBECO project?

• General model structure: Exploring spatiotemporal birch forest dynamics under influence from forestry in a homogeneous environment

• Adding realism I: Age-structured forest dynamics at scale 1 Ha

• Adding realism II: Dynamics in a heterogeneous landscape

• Adding realism III: Climate change scenarios• Adding realism IV: Influences from herbivores –

grazing ungulates and leaf-eating insects

0

500000

1000000

1500000

2000000

2500000

3000000

3500000

4000000

4500000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Topoclimatic models for the height gradient in the HIBECO model

8.0

10.0

12.0

14.0

16.0

18.0

20.0

0 100 200 300 400 500 600 700 800 900 1000

Height above bottom of valley, H

3-M

on

th M

ax. T

emp

erat

ure

, T

T = 15 +2 / (H/ 20 + 1) - 0.55 * H / 100

T = 13.7 -ABS(H-75)/80 + (1000+15.2*H)/2200

Purpose of the model

• An instrument linking various aspects of the HIBECO project closer together– HIBECO is an interdiciplinary project– Proper modeling requires explicit qualitative and quantitative

descriptions of essential variables and parameters and their dynamic interactions

• Exploring complex interactions in space and time: Scenarios– Basic scientific value– Practical value for management and politics: sustainable use of

birch forest resources and maintenance of the northern birch forest ecosystem

Sustainable use of mountain birch: How can modeling make a contribution?

• Birch is a keystone species in many parts of the Northern terrestrial ecosystems

• Mountain birch forest ranges are expanding due to a warmer climate and changes in cultural use of the landscape

• Mountain birch may become an increasingly important resource in the future (forestry etc.)

• Expansion of the forest ranges also means a parallel contraction of the treeless landscapes (problems related to reindeer herding and grazing field availability).

• Models may be used to explore long term interactions and consequences (“landscape behaviour”) under various hypothetical management regimes in virtual landscapes

Part 1: General Model Structure

Exploring spatiotemporal birch forest dynamics and disturbances in a

homogeneous environment

Pattern from intrinsic processes in a birch / forestry model system

100 m12.8 km

Rules for local cell disturbances (logging events)

• Rule 1:– At each time increment

randomly choose 5% of the cells

– Reset all chosen cells dominated by old-growth stands to young stands (i.e., local logging event at 1 Ha scale)

• Rule 2:– At each time increment

randomly choose one of the arena cells with old-growth stands

– Perform logging in this cell, and all connected cells with old-growth stands (i.e., logging with no spatial scale constraints up to arena size)

Time step: 10 12 14 16 18

Rule 1:

Rule 2:

Test: 150-year simulation

20 22 24 26 28

30 32 34 36 38

40 42 44 46 48

50 52 54 56 58

60 62 64 66 68

70 72 74 76 78

80 82 84 86 88

90 92 94 96 98

100 102 104 106 108

110 112 114 116 118

120 122 124 126 128

130 132 134 136 138

140 142 144 146 148

150

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

1 11 21 31 41 51 61 71 81 91 101 111 121 131 141

Time

"Old

gro

wth

" c

ells

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

"Dis

turb

ed

" c

ells

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

1 11 21 31 41 51 61 71 81 91 101 111 121 131 141

Time

"Old

gro

wth

" ce

lls

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

"Dis

turb

ed"

cells

Time series analysis of rule 2:

Large series (9200 time steps)

Fractal spatial pattern...

y = 4996x-0.8814

R2 = 0.9266

0

1000

2000

3000

4000

5000

6000

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Class

N

y = -0.881x + 3.6973

R2 = 0.9265

0

0.5

1

1.5

2

2.5

3

3.5

4

0 0.5 1 1.5 2 2.5 3 3.5 4

Log(class)

Lo

g(N

)

Disturbance size class histogram

y = -0.8435x + 2.6624

R2 = 0.7964

0

0.5

1

1.5

2

2.5

3

0 0.5 1 1.5 2 2.5 3 3.5 4

Log(class)

Lo

g(N

)

P(size class increment)/step)=0.9

n= ca 3700

Rule 2-pattern relatively insensitive to parameter adjustments

Here: relatively small local inter-cell variation in forest growth rate

Rule 2-pattern shows “mild chaos” on a return map of order 1

Comparing cover of old forest at time T (x-axis) with T+1 (y-axis)

y = -0.0003x2 + 2.1987x - 165.03R2 = 0.0566

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

Step T

Step

T+1

Polynomial:

Statistical mechanics: Self-organized critical state...

Landscape ecology: Shifting mosaic

Disturbance rules vs. landscape element connectivity

Rule 1 (scale-specific) Rule 2, less synchronized Rule 2, relatively synchronized

•Stop series at a random point where overall cover of old-growth cells is approximately 6500 cells

•For the next time step, set disturbance intensity to 0.05 (every 20th cell expected to be perturbed), and let all local disturbances spread to all connected old-growth cells (Rule 2)

•Result: Fractal patterns show better overall connectivity between similar growth phase elements, both for young and old age classes.

Disturbance rules vs. landscape element connectivity

Rule 1 (scale-specific) Rule 2, less synchronized Rule 2, relatively synchronized

All three landscape patterns above offer the same overall logging output in the long run.

Parameter-sensitive connectivity pattern, in particular “Rule 1”-based management

Local management areas define upper scale range for age class pattern even for “scale-free” logging rules

Grain: 1 Ha

Extent: 128*128 = 16,384 Ha

Extent: 9 * 16,384 Ha = 147,456 Ha

Extent: 400 * 147,456 Ha = 58,982,400 Ha

768

km

12.8

km

Part 2

Adding realism I: Age-structured forest dynamics at scale 1 Ha

Time

Example from a “Rule 2” type of logging regime

Intense computation:

For each time increment, update more than 100 year classes for each of the arena’s 16384 1 Ha pixels (cells)…

Time series for arena: Old growth forest cover (upper line) and logging intensity (bars)

0

500

1000

1500

2000

2500

30001 4 7

10

13

16

19

22

25

28

31

34

37

40

43

46

49

52

55

58

61

64

67

70

73

76

79

82

85

88

91

Time (years)

Nu

mb

er

of

1 H

a c

ells

0

100

200

300

400

500

600

700

800

1 4 7

10

13

16

19

22

25

28

31

34

37

40

43

46

49

52

55

58

61

64

67

70

73

76

79

82

85

88

91

Time series for one particular cell:

Num

ber

of o

ld tr

ees

Intra-cell dynamics

1

10

19

28

37

46

55

64

73

82

91

10

0

10

9

11

8

12

7

13

6

14

5

15

4

16

3

S1

S4

S7

S10

S13

S16

S19

0

10000

20000

30000

40000

50000

60000

70000

Number of trees

TimeAge class

Logging

Logging

1

10

19

28

37

46

55

64

73

82

91

10

0

10

9

11

8

12

7

13

6

14

5

15

4

16

3

S1

S5

S9

S13

S17

S21

0

2000000

4000000

6000000

8000000

10000000

12000000

Basal stem area

TimeAge class

Logging

Logging

Model behaviour needs careful validationExample: Time series for oldest age class (>=100 years) in a given 1 Ha cell

1 8

15 22

29 36

43 50

57

64

71

78

85

92

99

106

113

120

127

134

141

148

155

162

S1

0

100

200

300

400

500

600

Nu

mb

er o

f tr

ees

Time

Time series for oldest age class >100 years

Logging

Logging

Part 3

Adding realism II: Dynamics in a heterogeneous landscape

Adding GIS layers for local abiotic conditions

Relative growth rate smaller close to tree line:

A function of altitudal summer temperature gradient

Frequency histogram template for potential or actual birch forest substrate at 1 Ha scale from index 0 (no potential for birch) to 5 (full cover)

0

10

20

30

40

50

60

1 2 3 4 5

Index

Pe

rce

nt

100

m

Part 4

Adding realism III: Climate change scenarios

Topoclimatic models for the height gradient in the HIBECO model

8.0

10.0

12.0

14.0

16.0

18.0

20.0

0 100 200 300 400 500 600 700 800 900 1000

Height above bottom of valley, H

3-M

on

th M

ax. T

emp

erat

ure

, T

T = 15 +2 / (H/ 20 + 1) - 0.55 * H / 100

T = 13.7 -ABS(H-75)/80 + (1000+15.2*H)/2200

Series 1: No disturbances Series 2: “Rule 1”-disturbances Series 3: “Rule 2”-disturbances

Climate change scenarios:

+ 2.4 C over the next 120 years

(scenarios – not prognoses!)

Part 5:

Adding realism IV: Influences from herbivores – grazing ungulates and leaf-

eating insects

Grazing ungulates (sheep, reindeer) and forest structure

“Rule 1”-generated pattern “Rule 2”-generated patterns

Sustainable pluralistic utilization of birch forest:

Logging rules (“policy”) can significantly influence the overall quality and availability of forest elements for other uses, like grazing habitat for sheep and reindeer, hiking/tourism, etc.

Leaf-eating moth outbreaks (Epirrita, Operopthera etc.)

Lower limit: 100 m above bottom of valley

(Winter inversion)

Upper limit: 13.5 ºC isotherm for 3-month max summer temperature

Outbreak constraints:

In this preliminary and hypothetical 120-year scenario insects have tree-killing preudoperiodic outbreaks every 8-12 years

http://www.hibeco.org