Dynamics of natural temperate forests
- is there a universal model?
Kamil Král & Tomáš Vrška
in cooperation with David Janík, Pavel Daněk, Dušan Adam
The Silva Tarouca Research Institute, Department of Forest Ecology, Brno, Czech Republic
BlueCatresearchteamphoto:KateřinaSlámová
1994
2008
SALAJKA
(Beskydy)
1 2 3 4 5 6 7
What we are looking for?
To understand the
dynamics of tree layer in
space and time and the
spatial relations.
1994
2008
SALAJKA
(Beskydy)
1 2 3 4 5 6 7
What we are looking for?
Disturbances - events
making growing space
available (Picket et
White 1985; Oliver et
Larson 1990).
Disturbances:
- frequency
- distribution/range
- severity/intensity
- endo- or exogenous
Hypothese: Natural (primeval) temperate forest dynamics is a cyclical
process where the different patches with similar development are
cyclically changed in the time. The patches have different ratio of living
and dead wood and different developmental trends.
Questions (no hierarchic order):
- which features and variables should we measure?
- which scale of assessment should we use (how to assess
endogenous and exogenous disturbances)?
- how to separate and classify the parts of cycle with similar
processes (how to identify the patches of stages in the field/in the
datasets)?
Three steps:
 Definition and classification of stages and phases – 2008-2010
 Patch dynamics in the space and varibalitiy of patches on the
altitudinal vegetation gradient – 2011-2014
 Spatio(multi)-temporal dynamics – transition between stages and
phases – 2015-2017
Josef John – first idea how to described the dynamics
of temperate forests - 1851
Boubín - longest spatio-temporal dataset in the World
Šebková et al. 2011, Forest Ecology and Management
Beechwood
patch
phase
Leibundgut, 1959
Zukrigl, 1963 Mueller-Dombois, 1987
ukázky modelů vývojových cyklů temperátních lesů
Tabaku et al. 1999
Drössler et al. 2006
0
N
d 1,3
0
N
d 1,3
0
N
d 1,3
Developmental cycle model
(Korpel 1978, 1995)
timbervolume
m3
1000
500
250
0
750
0 200 300 400
years
100
phase of
expiration1st cycle
2nd cycle
3rd cycle
time
Stage of
Disintegration
Stage of
Growth
Stage of
Optimum
Stage of
Disintegration
phase of
regeneration
Stage of
Growth
phase of
expiration
(Korpel 1978,1995)
0
25
50
m
Kuuluvainen 2002
more US studies
more EU and JP studies
Which scale of assessment should we use?
Spatial scales and the often hierarchical nested occurence of different disturbance
factors
Kuuluvainen 2002
• limits of local and spatial
variability
• intra- and interspecific
competition
• single trees trajectory
Žofín – stem position map (1975-1997-2008)
live recruit
dead recruitno record
no record
no record
no record
no record
no record
no record
no record
no record – stem (still/already) doesn´t exist or doesn´t reach threshold d.b.h.
live recruits
2000s
1990s
1970s
Žofín: 2008
Žofín: 1997
Žofín: 1975
Different development in the time
Method of the moving filter – focal filtering
Live trees:
Dead trees:
d 1,3 [cm]
d 1,3 [cm]
(21m)
Mooving
Circle:
0
100
200
300
400
500
600
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
250
300
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
250
300
350
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
250
1-2 3-4 5-6 7-8 9-16
live
dead
Method of the moving filter – focal filtering
Live trees:
Dead trees:
d 1,3 [cm]
d 1,3 [cm]
(21m)
Mooving
Circle:
0
100
200
300
400
500
600
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
250
300
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
250
300
350
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
30
60
90
120
150
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
250
1-2 3-4 5-6 7-8 9-16
live
dead
0
50
100
150
200
12 34 56 78 910
live
dead
0
5
10
15
20
12 34 56 78 910
live
dead
0
50
100
150
12 34 56 78 910
live
dead
0
2
4
6
8
10
12 34 56 78 910
live
dead
0
50
100
150
200
12 34 56 78 910
live
dead
0
5
10
15
12 34 56 78 910
live
dead
0
20
40
60
80
100
12 34 56 78 910
live
dead
0
5
10
15
20
25
12 34 56 78 910
live
dead
N BA
Growth / expiration
Growth, initial
Growth, advanced
Steady State
DBH distributions
0
10
20
30
40
50
60
12 34 56 78 910
live
dead
0
5
10
15
20
12 34 56 78 910
live
dead
0
10
20
30
40
50
60
12 34 56 78 910
live
dead
0
5
10
15
20
25
12 34 56 78 910
live
dead
0
10
20
30
40
50
12 34 56 78 910
live
dead
0
5
10
15
20
12 34 56 78 910
live
dead
0
20
40
60
80
100
12 34 56 78 910
live
dead
0
5
10
15
20
12 34 56 78 910
live
dead
N BA
Optimum, typical
Optimum, ageing
Breakdown, initial
Breakdown / regeneration
DBH distributions
Classification using Artificial Neural Network
Live trees:
Dead trees:
d 1,3 [cm]
d 1,3 [cm]
Stage:
Growth
Optimum
Breakdown
Steady State
Resulting map of patches - developmental stages and phases
Legend:
(6%)
(16%)
(20%)
(9%)
(14%)
(21%)
(15%)
STAGE
Portion
of Area
Growth 21%
Optimum 29%
Disintegration 35%
Max. stability 15%
TOTAL 100%
STEADY STATE
Steady state
Accuracy of Artificial Neural Network classification
Král,K., Vrška,T., Hort,L., Adam,D., Šamonil,P., 2010: Developmental phases in a temperate natural spruce-firbeech
forest: determination by a supervised classification method. European Journal of Forest Research 129,
339-351.
86 % 91 %
Growth / expiration
Growth, initial
Growth, advanced
Optimum, typical
Optimum, ageing
Breakdown, initial
Breakdown / regeneration
Steady State
Growth
Optimum
Breakdown
Steady State
PHASE STAGE
0
N
DBH
0
N
DBH
0
N
DBH
Steady state
Optimum
Growth
Breakdown
N
0 DBH
living trees
dead trees
0
N
DBH
0
N
DBH
0
N
DBH
Steady state
Optimum
Growth
Breakdown
N
0 DBH
living trees
dead trees
Král,K., Vrška,T., Hort,L., Adam,D., Šamonil,P., 2010: Developmental phases in a temperate natural spruce-firbeech
forest: determination by a supervised classification method. European Journal of Forest Research 129,
339-351.
Three steps:
 Definition and classification of stages and phases – 2008-2010
 Patch dynamics in the space and varibalitiy of patches on the
altitudinal vegetation gradient – 2011-2013
 Spatio(multi)-temporal dynamics – transition between stages and
phases – 2014-2015
Study Site
Census
area
[ha]
Altitude min.
[m a.s.l.]
Altitude
max. [m
a.s.l.]
Mean
annual
temp. [°C]
Mean annual
prec. totals
[mm]
Years of
census
Cahnov 17.3 150 153 9.3 517 73’, 94’, 06’
Ranšpurk 22.3 152 155 9.3 517 73’, 94’, 06’
Salajka 19.0 715 815 5.4 1144 74’, 94’, 07’
Žofin 74.5 735 835 4.3 866 75’, 97’, 08’
Boubín 46.7 910 1110 4.0 867 72’, 96’, 10’
0 50 100 150 20025
m
±
Legend:
Growth
Optimum
Breakdown
Steady State
1974 1994 2007
Salajka
• European beech > 80%
• Silver fir and Norway spruce < 10% each
• Altitude: 715 - 815 m a.s.l.
• Strictly protected since 1937; 19 ha
silver fir
dieback
silver fir
dieback
0 100 200 300 40050
m
1975
Legend:
Growth
Optimum
Breakdown
Steady State
1997 2008
±
• European beech 65%
• Norway spruce 33% and silver fir < 2%
• Altitude: 735 - 830 m a.s.l.
• Strictly protected since 1838; 72 ha !
2007
Kyrill
Žofín
1972
Legend:
Growth
Optimum
Breakdown
Steady State
1996 2010
±
Boubín
• European beech 54%
• Norway spruce 44% and silver fir < 2%
• Altitude: 930 - 1110 m a.s.l.
• Strictly protected
since 1858
• 45 ha
2008
Emma
Stage Proportion [%]
0%
10%
20%
30%
40%
50%
1974 1994 2007 1975 1997 2008 1972 1996 2010
Salajka Žofín Boubín
Growth Optimum Breakdown Steady StateMean Patch Size [m]
0
500
1000
1500
2000
2500
3000
1974 1994 2007 1975 1997 2008 1972 1996 2010
Salajka Žofín Boubín
Growth Optimum Breakdown Steady State Total
• The proportion of stages varies
among sites and also in time
• Growth stage cover usually
30 – 40 %
• Breakdown stage is usually
10-20 %
• Steady State seems to increase
along altitude (18 -> 38 %)
• The MPS is usually higher than
average for the Growth stage
(in Boubin alternated by SS)
• MPS is always subnormal for
Breakdown stage
• Mean Patch Size is even at
the level of the whole mosaic!
Patch Analyst 5.1
KRÁL K., McMAHON S.M., JANIK D., ADAM D., VRŠKA T., 2014: Patch mosaic of developmental stages in central European
natural forests along vegetation gradient. Forest Ecology and Management 330: 17–28.
In contrast to earlier hypotheses, it turns out the patch dynamics has the similar
parameters in the N-E US forests:
KRÁL K., SHUE J., VRŠKA T., GONZALES-AKRE E.B., PARKER G.G.,
McSHEA W.J., McMAHON S.M., 2016. Fine-scale patch mosaic of
developmental stages in Northeast American secondary temperate forests: the
European perspective. European Journal of Forest Research 135 (5): 981-996.
Three steps:
 Definition and classification of stages and phases – 2008-2010
 Patch dynamics in the space and varibalitiy of patches on the
altitudinal vegetation gradient – 2011-2013
 Spatio(multi)-temporal dynamics – transition between stages and
phases – 2014-2015
Christensen et al. 2007
Christensen et al. 2007
0 50 100 150 20025
m
±
Legend:
Growth
Optimum
Breakdown
Steady State
1974
1994
2007
Multi-temporal comparisons – transitions between stages and phases
Rule-based classification of developmental stages and phases
- ArcGIS Toolbox
- the DBH bins used in different forest types were defined and justified in Král et al. (2014a)
- 10 developmental phases described and portrayed by respective local DBH distributions
characteristic for individual developmental phases
Empiric classification of transitions
- all phase-to-phase transitions were quantified between the 70’s and 90’s, 90’s and 00’s and
70’s and 00’s.
- descriptive categories:
- Stable – the developmental phase remained unchanged between censuses
- Progressive – the phase was shifted forward in the cycle
- Regressive – the phase was shifted backward in the cycle
- Disturbance – a shortcut to early developmental phases likely caused by a disturbance
- No trend – development with no clear direction along the forest cycle
- Unlikely – unlikely development (a possible misclassification of the phase in either of the
observations).
Quantitative evaluation of transitions
- 10,000 bootstrap samples we derived null distributions for all transition frequencies
- for each research plot we used a bootstrap sample size equal to the number of nonoverlapping
moving windows necessary to cover the whole area of the plot
- we used a sequential Bonferroni-type procedure (Benjamini & Hochberg 1995), which
controls for a false discovery rate.
example Boubín
Green – higher number of transitions than the null model
Red – lower number of transitions than the null model
The proportion of transitions following preferential, randomly frequent
and uncommon pathways
Research plots:
Boubí
n
Žofín
Salajk
a
Cahno
v -
Ranšp
urk
Mean SEM
90s-00s
Period (years) 14 11 13 12 12.5 0.6
Preferential pathways (%) 70.4% 66.1% 70.2% 63.0% 67.4 1.8
Randomly frequent pathways (%) 21.1% 18.8% 27.8% 29.5% 24.3 2.6
Uncommon pathways (%) 8.5% 15.1% 2.0% 7.5% 8.3 2.7
70s-90s
Period (years) 24 22 20 21 21.8 0.9
Preferential pathways (%) 76.5% 75.1% 49.6% 57.6% 64.7 6.6
Randomly frequent pathways (%) 13.5% 12.5% 48.4% 37.7% 28.0 9.0
Uncommon pathways (%) 10.0% 12.4% 2.0% 4.6% 7.2 2.4
70s-00s
Period (years) 38 33 33 33 34.3 1.3
Preferential pathways (%) 69.7% 61.8% 60.9% 59.7% 63.0 2.3
Randomly frequent pathways (%) 23.0% 28.3% 38.3% 35.5% 31.3 3.5
Uncommon pathways (%) 7.3% 9.8% 0.9% 4.8% 5.7 1.9
The proportion of the three major transition categories in all observations
Research plots: Boubín Žofín Salajka
Cahnov -
Ranšpurk
Mean SEM
90s-00s
Period (years) 14 11 13 12 12.5 0.6
No transitions (%) 46.6% 45.0% 46.0% 51.7% 47.4 1.5
Cyclic transitions (%) 14.7% 22.0% 27.4% 15.7% 20.0 3.0
Acyclic transitions (%) 38.6% 33.0% 26.5% 32.5% 32.7 2.5
70s-90s
Period (years) 24 22 20 21 21.8 0.9
No transitions (%) 30.0% 37.4% 26.2% 34.2% 31.9 2.5
Cyclic transitions (%) 16.7% 23.4% 37.0% 27.8% 26.2 4.3
Acyclic transitions (%) 53.4% 39.2% 36.8% 38.0% 41.8 3.9
70s-00s
Period (years) 38 33 33 33 34.3 1.3
No transitions (%) 19.2% 22.5% 20.5% 22.2% 21.1 0.8
Cyclic transitions (%) 20.1% 31.0% 40.8% 34.8% 31.7 4.4
Acyclic transitions (%)
60.8% 46.4% 38.6% 43.0% 47.2 4.8
Three main outputs
• in total about 65% of all observed phase-to-phase transitions were significantly
more frequent than random switches between phases
• about 28% of observed transitions proceeded along pathways of random
frequency
• only about 7% of observed transitions were realized through pathways
significantly less frequent than random switches between phases
• the mean ratio of cyclic/acyclic transitions (2:3) was more or less stable
throughout time
• in average only less than 40% of transitions between different developmental
phases were classified as cyclic (following the model cycle), the majority of
these transitions were realized through significantly frequent preferential
pathways
Král K., Daněk P., Janík D., Krůček M. & Vrška T., 2017. How cyclical and
predictable are central European temperate forest dynamics in terms of
developmental phases? Journal of Vegetation Science – online first
Multitemporální analýza
%
%
%
%
%
DĚKUJI OTÁZKY