Potential of temporal and spatial analysis techniques as tools for exploring changes in forest site productivity and following environmental interactions [Elektronische Ressource] : Finnish-German comparative study / vorgelegt von Pedro José Pérez Martínez

De
Potential of temporal and spatial analysis techniques as tools for exploring changes in forest site productivity and following environmental interactions. Finnish-German comparative study Inaugural-Dissertation zur Erlangung der Doktorwürde der Forst- und Umweltwissenschaften Fakultät der Albert-Ludwigs-Universität Freiburg i. Brsg. vorgelegt von Pedro José Pérez Martínez geboren in Barcelona Freiburg im Breisgau 2003 Dekan: Pr. Dr. Karl-Reinhard Volz Referent: Pr. Dr. Heinrich Spiecker Korreferent: Pr. Dr. Jürgen Bauhus A mis padres, Perennidad de los ríos. Conveniencia de los montes …la hermosura provechosa de los montes, firmes costillas del cuerpo muelle de la tierra, aumentando su hermosa variedad: en ellos se recogen los tesoros de las nieves, se forjan los metales, se detienen las nubes, se originan las fuentes, anidan las fieras, se empinan los árboles para las naves y edificios y donde se guarecen las gentes de las avenidas de los ríos, se fortalecen contra los enemigos y gozan de salud y vida. (Baltasar Gracián 1651-1657, “El Criticón”) CONTENTS I CONTENTS Abstract .........................................................................................................
Publié le : jeudi 1 janvier 2004
Lecture(s) : 18
Source : FREIDOK.UB.UNI-FREIBURG.DE/FREIDOK/VOLLTEXTE/2004/1361/PDF/TTDEFINITIVO.PDF
Nombre de pages : 206
Voir plus Voir moins










Potential of temporal and spatial analysis techniques as
tools for exploring changes in forest site productivity
and following environmental interactions.
Finnish-German comparative study








Inaugural-Dissertation zur
Erlangung der Doktorwürde
der Forst- und Umweltwissenschaften Fakultät
der Albert-Ludwigs-Universität
Freiburg i. Brsg.



vorgelegt von



Pedro José Pérez Martínez
geboren in Barcelona



Freiburg im Breisgau
2003






































Dekan: Pr. Dr. Karl-Reinhard Volz

Referent: Pr. Dr. Heinrich Spiecker
Korreferent: Pr. Dr. Jürgen Bauhus



A mis padres,





















Perennidad de los ríos. Conveniencia de los montes

…la hermosura provechosa de los montes, firmes costillas del cuerpo muelle de la tierra,
aumentando su hermosa variedad: en ellos se recogen los tesoros de las nieves, se forjan los
metales, se detienen las nubes, se originan las fuentes, anidan las fieras, se empinan los
árboles para las naves y edificios y donde se guarecen las gentes de las avenidas de los ríos,
se fortalecen contra los enemigos y gozan de salud y vida.

(Baltasar Gracián 1651-1657, “El Criticón”)


CONTENTS I
CONTENTS

Abstract ...............................................................................................................................1
zusammenfassung................................................................................................................4
Sumario8
1 Introduction..............................................................................................................11
1.1 Problem to be solved ........................................................................................13
1.2 Investigated potential causes ............................................................................14
2 Material ....................................................................................................................16
2.1 Environmental data...........................................................................................16
2.2 Height Stem analysis data ................................................................................17
Sampling strategy of the Level II sites in Finland and Germany .................19
3 Methods41
3.1 Radial increment variation and chronology building .......................................41
3.2 Climate-growth relationships ...........................................................................45
3.3 Radial increment growth trends, Anomalies and climate.................................45
3.4 Height growth anomalies and trends ................................................................46
3.4.1 Height growth anomalies and methodological limitations ...........................47
3.4.2 Yield tables...................................................................................................48
3.4.3 Height growth trends....................................................................................51
3.4.4 Analysis of long-term trends using deterministic potential growth models,
growth models incorporating indices of annual climatic variation and physiological
process based models52
3.5 examining the Relationships between changes of growth and nutrition by the
use of Generalized linear and regression models .............................................................56
3.6 examining the Relationships between changes of growth and nutrition by the
use of Spatial interpolation methods ................................................................................58
3.6.1 Variography: the structure identification......................................................58
3.6.2 The cross validation of the model ................................................................60
3.6.3 Kriging with external drift............................................................................61
4 Results......................................................................................................................63
4.1 Radial growth chronologies and Growth responses to climate.........................63
4.2 Growth trends and relationships.......................................................................81
4.2.1 Recent trends in the radial growth and climate in Germany81
4.2.2 Height and radial growth in relation to levels of recognized causal factors in
Eastern Germany..........................................................................................................86
4.2.3 Temporal and spatial trends in height growth changes and relation to levels
of putative causal factors..............................................................................................91
4.2.4 Long-term trends in height increment variation for expressing growth
anomalies using deterministic potential growth models.............................................121
4.2.5 Long-term trends in height increment variation for expressing growth
anomalies using growth models without potentials....................................................125 II Pérez-Martínez
4.2.6 Long-term trends in height increment variation for expressing growth
anomalies using physiological process based models. Growth of trees using the
Finnfor model.............................................................................................................130
4.2.7 Results of liming, nitrogen and phosphorous fertilization effect on tree-rings
and shoot length of a Norway spruce trees on a long-term permanent experiment in
South Western Germany ............................................................................................133
4.3 Estimation with external drift.........................................................................136
Interpolation and mapping..........................................................................140
5 Discussion ..............................................................................................................142
The adequacy of using yield tables ................................................................146
6 Conclusion151
Background and Acknowledgements ..............................................................................156
Annex 1 ...........................................................................................................................157
Determination of the degree of the drift ......................................................................157
The inference of the Optimal Generalized Covariance ...............................................158
Annex 2161
Kriging with external drift...........................................................................................161
Literature Cited ...............................................................................................................164

























CONTENTS III
LIST OF TABLES

aTable 1 Descriptive summary statistics of the sites included in the study by the two
countries for pine and spruce......................................................................................21
Table 2 Location, elevation, and characteristics of the sites included in the study ............22
Table 3 Values of the three main variables, Temperature change (K), Precipitation change
(%) and Foliar nitrogen (mg/g), grouped by treatment classes. The values were
classified into classes according to the distribution (25th and 75th percentiles for low
and high value classes respectively) of the three main variables over the Level II
dataset along Europe (n=195 plots of Norway spruce, n=216 plots of Scots pine). *A
special class was built for stands growing under boreal conditions and for the
Temperature parameter...............................................................................................36
Table 4 Selected sites form the level-II network in Germany and Finland for Scots pine
and Norway spruce according to levels in Temperature change (K), Precipitation
change (%) and foliar nitrogen (mg/g) .......................................................................37
Table 5 Suitable top height site index curves used as references for the sampling sites
where only one generation of trees was collected.......................................................49
Table 6 Descriptive statistics of standard radial increment chronologies for samples from
Finnish study sites for the period 1873-2000. Mean sensitivity, standard deviation,
and autocorrelation values are for standard chronologies (Cook, 1985).
Standardization method: Double detrending, a rigid spline (with 50% frequency cut-
off in 75 % of the series length) was fitted to a ring-width series; a more flexible
spline (with 50% frequency cut-off in 10 % of the series length) was fitted to the ring-
width-index derived from the first detrending ............................................................67
Table 7 Descriptive statistics of residual radial increment chronologies for samples from
Finnish study sites for the period 1873-2000. Mean sensitivity, standard deviation,
and autocorrelation values are for residual chronologies (Cook, 1985).
Standardization method: Double detrending. Autocorrelation was removed in each
series using autoregressive moving average (ARIMA) time series model (the order
was selected for the individual series by searching the first minimum of the Akaike
Information Criterion). Other definitions from Table 6 .............................................69
Table 8 Descriptive statistics of standard radial increment chronologies for samples from
German study sites for the period 1812-2000. Mean sensitivity, standard deviation,
and autocorrelation values are for standard chronologies (Cook, 1985).
Standardization method: Double detrending, a rigid spline (with 50% frequency cut-
off in 75 % of the series length) was fitted to a ring-width series; a more flexible
spline (with 50% frequency cut-off in 10 % of the series length) was fitted to the ring-
width-index derived from the first detrending. ...........................................................70
Table 9 Descriptive statistics of residual radial increment chronologies for samples from
German study sites for the period 1812-2000. Mean sensitivity, standard deviation,
and autocorrelation values are for residual chronologies (Cook, 1985).
Standardization method: Double detrending. Autocorrelation was removed in each
series using autoregressive moving average (ARIMA) time series model (the order
was selected for the individual series by searching the first minimum of the Akaike
Information Criterion). ...............................................................................................71
Table 10 Statistics of tree radial increment data standard and residual chronologies of all
Finnish and German sites grouped by tree species (Norway spruce and Scots pine);
values: means for the two chronology types (standard and residual) with range of IV Pérez-Martínez
values (minimum and maximum), calculations based on group-wise common overlap
periods ........................................................................................................................72
Table 11 Cross-correlation coefficients (r ) and corresponding p-values among radial xy
increment residual chronologies along Finland. Calculations based on maximum
common overlap period (1956-2000, n=45) ...............................................................74
Table 12 Cross-cor ) and corresponding p-values among radial xy
increment residual chronologies along Germany. Calculations based on maximum
common pair wise overlap period (1934-1999). Only significant correlations are
shown (p<0.01)...........................................................................................................75
Table 13 Statistics of the tree ring data of the Pfalzgrafenweiler site Fi-337 based on a
total of 68 individual tree series. Averages of mean sensitivity (MS), standard
deviation (sd), first order serial autocorrelation [AR(1)], signal to noise ratio (SNR),
variance of first eigenvector , mean correlation between trees and Pearson correlation
of sample trees............................................................................................................82
Table 14 Continuous and categorical variables, concerning climatic, deposition, nutrition
and combined factors, assessed and used as input variables in the GLM...................93
Table 15 Time trend in Height growth deviation (Y = a + b*DATE). As well as time trend
in height growth deviation, simple linear regressions were calculated for other
explanatory variables. The results are showing the magnitude of the detected changed
given by the slopes of the calculated trend lines. Linear regression equations were
calculated for each individual site. Only Finnish sites included in the computations.95
Table 16 Final regression equation (SAS multiple regression and ANOVA procedure:
linear model) used to predict height growth deviations for Scots pine and Norway
spruce in Finland using indicator variables (basically climate and deposition). Tree
specie included as a dummy variable. Estimates for height growth deviations on
Norway spruce potential site are reduced by –1.887 %. Only Finnish sites included
in the computations.....................................................................................................96
Table 17 The correlations table displays Pearson correlation coefficients, significance
values (2-tailed), and the number of cases with non-missing values (31 sample data-
points at Finnish sites). Sd_Ih: slope dependent of the height growth deviation (%
-1 2year ). Fi98 : bulk open nitrogen oxidized deposition for the year 1998 (mg N/m ). _nox
2Fi98 : bulk open nitrogen reduced deposition for the year 1998 (m ). _nrd
NDVI : normalized vegetation index during the growing season (May-_MaySep_89
September) and for the year 1989 (expressed on a scale from -1 to 1). Si : slope _nred
2 -1independent of the reduced nitrogen deposition (mg N/m year ) .............................97
Table 18 Height growth model fitting statistics (parameter estimates and residual errors by
non-linear regression) for Scots pine (Level II 15/09) and Norway spruce (Level II
15/17) Finnish plots..................................................................................................122
Table 19 Height growth model fitting statistics (parameter estimates and residual errors by
non-linear regression) for Scots pine (Level II 15/09) old and young generation and
for the basic and the complete Schumacher models which include temperature
climatic index in both the asymptote and age function.............................................126
Table 20 Change in predictive potential productivity for trees of two cohorts growing
under altered climatic conditions in relation to a normal climatic condition. Potential
productivity using a site index approach (height a crop achieves at a given age, h : 50
average tree height at age in stock height t = 50 years, h : average tree height at 0.2 100
age in stock height t =100 years) and Schumacher equations A-B. ∆ h : site index 0.2 50
deviation (%) at the reference age of 50 years and within each climate scenario.
∆ h : site index deviation (%) at the reference age of 100 years and within each 100CONTENTS V
climate scenario (in both cases I used the site index of the older as expected value of
the younger cohort in scenario wise). h dev (%) : height growth deviation of the 100
two altered climate scenarios (temperature increase, temperature decrease) respect the
control scenario (current seasonal air temperature) within each generation. The
consistent results of both models indicate that different cohorts have individualistic
growth responses to future climatic change (PAN & RAYNAL 1995)....................127
Table 21 Characteristics of the site included in the simulations (2000) ............................132
Table 22 Data on fertilizer (lime, phosphor and nitrogen) application.............................133
Table 23 Ring-width at breast height and shoot length increment in fertilization treatment
combinations, 1953-1974. Values are means of composite ring widths of n = 5-10
trees per plot with SEs given in parentheses. Treatment values within a column with
different letters are significantly different at α=0.05 (Bonferroni test).....................134
Table 24 Model parameters for height increment deviation.............................................138































VI Pérez-Martínez
FIGURE CAPTIONS

Figure 1 Map of the study area. Contour curves represent deposition of oxidized nitrogen
2(NOx (mg N/m )) during the year 1998 (Co-operative Programme for Monitoring and
Evaluation of the Long-Range Transmission of Air Pollutants in Europe: EMEP
source). .......................................................................................................................26
Figure 2 Air temperatures (left) and precipitation (right) at the selected sites (below
Germany, upper Finland). The solid horizontal line is the average value for this
hundred year period. Temperatures are monthly averages and precipitations are
monthly sums for the vegetation period (May to September). This dataset was part of
the CRU (Climatic Research Unit, University of East Anglia). .................................28
Figure 3 Stem analysis for reconstructing annual radial and height increment. At different
heights stem discs were sampled and radial increment measured in eight cardinal
directions. In the field annual shoot increments were taken. Comparing the number of
shoots with the number of tree rings at a given stem height, quality of measurements
is controlled (GERECKE 1988). ................................................................................29
Figure 4 Annual radial increment (mm) vs calendar year from the individual trees sampled
at the Finnish long-term fertilization METLA sites. Each box represents a plot and
each line represents a disc in 1.3 m. Note the generally high level of conformity
between trees at each site. The codes of the sites are from table 2. ............................30
Figure 5 Annual radial increment (mm) vs calendar year from the individual trees sampled
at the Finnish Level II sites. Each box represents a plot and each line represents a disc
in 1.3 m. Note the generally high level of conformity between trees at each site. The
codes of the sites are from table 2...............................................................................31
Figure 6 Annual radial increment (mm) vs calendar year of Norway spruce sampled at 14
plots in Germany. Each box represents a plot and each line represents a disc in 1.3 m.
Note the generally high level of conformity between trees at each site. The codes of
the sites are from table 2. ...........................................................................................32
Figure 7 Annual radial increment (mm) vs calendar year of Scots pine sampled at 17 plots
in Germany. Each box represents a plot and each line represents a disc in 1.3 m. Note
the generally high level of conformity between trees at each site. The codes of the
sites are from table 2. .................................................................................................33
Figure 8 Tree-ring mean chronologies from some Finnish and German sampling sites and
trees (n=5) that were among the dominant trees in the past and presently.
Chronologies were built by averaging all the five trees in the plots. The codes of the
sites are from table 2.35
Figure 9 Selected sites form the level-II network in Germany and Finland for Scots pine
and Norway spruce according to levels in Temperature change (K), Precipitation (%)
and Foliar nitrogen (mg/g). Tmpdif3159 (K): change in baseline seasonal air
temperature (absolute difference in long-term mean seasonal (May-Sep) air
temperature (in K) between two reference periods 1961-1990 and 1901-1930.
Nsdif3159p (%): change in baseline seasonal precipitation (deviation of the long-term
mean seasonal (May-Sep) precipitation sum of the reference period 1961-1990 in
percent of the reference period 1901-1930. N_0avg (mg/g), foliage N-nutrition
(average N concentration of the annual foliage measured over the period 1993-1998.
....................................................................................................................................39
Figure 10 Site tree ring chronologies and site index chronologies (upper graphs Norway
spruce and Scots pine German sites, lower graphs Norway spruce and Scots pine
Finnish sites)...............................................................................................................44

Soyez le premier à déposer un commentaire !

17/1000 caractères maximum.