Long-term changes in tree species composition in old-growth dinaric beech-fir forest

Abstract

Successive inventories of individual old-growth beech–fir forests from Dinaric mountains show structural changes during the last half a century. To be aware of these trends is important for general understanding of regeneration processes in old-growth forests, as well as for stating appropriate silvicultural aims. However, no study comparatively investigated data sets from several oldgrowth forests. We analyzed data from 31 inventories of growing stock in seven old-growth forests representing large area stretching from the north-west of the Dinaric mountain range in Slovenia to the central part in Croatia. All forests
experienced a decrease in the proportion of silver fir in the growing stock. The decrease was of a different magnitude in individual old-growth forests. However, the overall growing stock in the majority of old-growth forests didn’t fluctuate as expected due to the silver fir decline. The diameter distribution of five selected forests was closer to a rotated sigmoid than to a negative exponential distribution in spite of relatively large areas which were inventoried. The curves for growing stock according to dbh revealed, that beech and silver fir occupy different niche within ecosystem. In the lower diameters beech is more frequent
and competitive, while silver fir compensates for this in larger diameters (and heights).Both curves from most reserves were not stable in time. The most apparent and worrying trend in Slovenia was the decrease of the silver fir curve over the entire range of the diameter distribution, with peaks in the lowest and largest diameters. Moreover, the silver fir regeneration > 20 cm was almost completely absent due to high densities of large herbivores. While Corkova Uvala appears to be a very balanced reserve in regard to all sampled parameters. The results revealed a general synchronous trend of silver fir replacement by beech on a broader geographical scale. The complex reasons for this are discussed and some general guidelines for the silviculture in managed beech fir forests are given.

Full text

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
13
UDK: 630*228.81+535 Original scientific paper
Izvorni znanstveni ~lanak
LONG-TERM CHANGES IN TREE SPECIES
COMPOSITION IN OLD-GROWTH DINARIC
BEECH-FIR FOREST
DUGORO^NE PROMJENE U SASTAVU VRSTA DRVE]A DINARSKIH
BUKOVO-JELOVIH PRA[UMA
JURIJ DIACI1, DU[AN RO@ENBERGAR1, STJEPAN MIKAC2, IGOR ANI]2,
TOMA@ HARTMAN3, ANDREJ BON^INA1
1 University of Ljubljana, Biotechnical Faculty, Department of Forestry and
Renewable Forest Resources, Ve~na pot 83, 1000 Ljubljana, Slovenia
2 University of Zagreb, Faculty of Forestry, Croatia
3 Slovenia Forest Service
Received – Prispjelo: 15. 10. 2007. Accepted – Prihva}eno: 14. 3. 2008.
Successive inventories of individual old-growth beech–fir forests from Di-
naric mountains show structural changes during the last half a century. To be
aware of these trends is important for general understanding of regeneration
processes in old-growth forests, as well as for stating appropriate silvicultural
aims. However, no study comparatively investigated data sets from several old-
growth forests. We analyzed data from 31 inventories of growing stock in seven
old-growth forests representing large area stretching from the north-west of the
Dinaric mountain range in Slovenia to the central part in Croatia. All forests
experienced a decrease in the proportion of silver fir in the growing stock. The
decrease was of a different magnitude in individual old-growth forests. Howe-
ver, the overall growing stock in the majority of old-growth forests didn’t fluc-
tuate as expected due to the silver fir decline. The diameter distribution of five
selected forests was closer to a rotated sigmoid than to a negative exponential
distribution in spite of relatively large areas which were inventoried. The curves
for growing stock according to dbh revealed, that beech and silver fir occupy
different niche within ecosystem. In the lower diameters beech is more frequent
and competitive, while silver fir compensates for this in larger diameters (and
heights).Both curves from most reserves were not stable in time. The most appa-
rent and worrying trend in Slovenia was the decrease of the silver fir curve over
the entire range of the diameter distribution, with peaks in the lowest and largest

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
14
diameters. Moreover, the silver fir regeneration > 20 cm was almost completely
absent due to high densities of large herbivores. While ^orkova Uvala appears
to be a very balanced reserve in regard to all sampled parameters. The results
revealed a general synchronous trend of silver fir replacement by beech on a
broader geographical scale. The complex reasons for this are discussed and some
general guidelines for the silviculture in managed beech fir forests are given.
Keywords: old-growth forest, beech-fir forest, virgin forest, diameter distribu-
tion, growing stock, species composition, alternation of tree spe-
cies, long-term inventories.
INTRODUCTION
UVOD
Beech-fir forests form one of the largest areas of continuous forest in south-
central Europe. In the north-western part of the Balkan Peninsula, they stretch over
the Dinaric Mountain range along the Adriatic Sea coast, covering some 163.500
ha in Slovenia and 140.000 ha in Croatia. They serve as both an important wood
source and a key habitat for several important and endangered animal species. Du-
ring the last century the tree species composition of this forest changed dramatically,
especially in Slovenia (Ficko in Boncina, 2006), where once silver fir (Abies alba
Mill.) dominated forest is now in some areas dominated by beech (Fagus sylvatica
L.). The mechanisms behind the alternation in dominance that occurred over the
past century are not entirely clear. There are several possible reasons, including
reintroduction of red deer in the late 19th century, silver fir decline, which started
in the 50’s (Bon~ina et al. 2003), different management regimes that changed the
stand climate in favor of beech and natural processes of tree species replacement
(Brinar 1969; Forcier 1975; Fox 1977; Ga{per{i~ 1974; Mlin{ek 1967). The silver
fir decline in old-growth forests suggests that management was not the predominant
factor triggering it. Although repetitive studies of structural changes of single old-
growth forests from Dinaric mountains are frequent (e.g. (Hartman 1987; Turk et al.
1985; Ro`enbergar 2000), there is no overview of structural changes in old-growth
forest from several sites. In this study long-term data about structure and tree species
composition of mature stands and regeneration was analyzed from several selected
old-growth forest reserves in the area of Dinaric beech–fir forests in Slovenia and
Croatia. The aims of the study were to: (1) examine if virgin forest from different
geographical regions of Dinaric mountain chain show similar trends of structural
changes over the time, (2) propose ecological reasoning for the changes, (3) identify
possible future trends of forest structures and (4) develop general recommendations
for managed forests.

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
15
METHODS
METODE
Research sites
Podru~je istra`ivanja
The research was performed in 7 natural old-growth forest reserves. Forest re-
serves Rajhenavski Rog (RR), Pe~ka (PE), Strmec (ST), Krokar (KR) and Bukov vrh
(BV) are located in Slovenia, while ^orkova uvala (CU) and Dev~i}a tavani (DT) in
Croatia (Figure 1). In all cases the site conditions were similar (Table 1) and forests
were dominated by beech - fir (Fagus sylvatica L. and Abies alba Mill.) communi-
ties, which are typically located between 700-1200 meters in the Dinaric mountain
range. Other less abundant species were also present in forest stands, including syca-
more maple (Acer pseudoplatanus L.), wych elm (Ulmus glabra Huds.), spruce (Picea
abies (L.) Karsten), common ash (Fraxinus excelsior L.), and large-leaved lime (Tilia
platyphyllos Scop.).
RR
CU
DT
ST
BV
KR
PE
Austria
Bosna and Herzegovina
Hungary
CROATIA
SLOVENIA
Italy
Adriatic sea
Figure 1: Locations of the old-growth beech–fir forest reserves included in the research (Rajhenavski
Rog – RR, Pe~ka – PE, Strmec – ST, Krokar – KR, Bukov vrh – BV, ^orkova uvala – CU and
Dev~i}a tavani – DT)
Slika 1. Bukovo-jelove pra{ume uklju~ene u istra`ivanje (Rajhenavski Rog – RR, Pe~ka – PE, Strmec
– ST, Krokar – KR, Bukov vrh – BV, ^orkova uvala – CU i Dev~i}a tavani – DT)
Two of the selected research sites, namely Krokar and Bukov vrh, are located
on the edge of beech–fir forest distribution area. The conditions are not as favorable
for silver fir growth as in other cases; therefore the proportion of silver fir on these
locations is by nature lower compared to other analyzed forest reserves (Kordi{,
1985). Krokar is situated on a plateau just above the Kolpa river with more intensive

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
16
exposure to warm air which is lifting up along the cliffs above the river. The large
area of the reserve is also covered with dolomite parent material changing to some
extend the soil conditions to more uniform which promotes beech as dominant tree
species (Zeibig, 2001). Bukov vrh on the other hand is close to altitude border of
silver fir distribution with colder mountain climate.
Typically Dinaric Mountains consist of dolomite and limestone parent material;
therefore the soils are in all cases free draining, ranging from rendzinas to calcareous
brown soils (calcocambisol) of variable depths. Macro- and microtopography of the
research sites are diverse, changing the site conditions significantly over very small
spatial scales. Karst phenomena such as sinkholes and rock outcrops at or close to
the surface are common on all sites.
Table 1. Basic data of the research sites (RR – Rajhenavski rog, PE – Pe~ka, KR – Krokar, ST – Strmec,
BV – Bukov vrh, DT – Dev~i}a tavani, CU – ^orkova uvala)
Tablica 1. Osnovni podaci o istra`ivanim lokalitetima (RR – Rajhenavski rog, PE – Pe~ka, KR – Krokar,
ST – Strmec, BV – Bukov vrh, DT – Dev~i}a tavani, CU – ^orkova uvala)
Research site / Lokalitet RR PE KR ST BV DT CU
Area / Povr{ina [ha]52.1 59.5 74.5 15.6 9.3 100 79.5
Location
Polo`aj 45°40’N
15°01’E 45°46’N
15°00’E 45°33’N
14°47’E 45°38’N
14°49’E 46°00’N
13°53’E 44°88’N
15°04’E 44°55’N
15°32’E
Altitude
Nadm. vis.740-880 795-910 840-1170 840-940 1200-
1313 1192-
1295 860-1030
Annual precipitation
God. padal. [mm]
∼1650 ∼1220 ∼1526 ∼1556 ∼3000 1875 ∼1650
Average annual
temperature
Prosj. god. temp. [°C]
7 14.3 8.4 8.3 6.2 5 7
Meteorological statiom
Meteor. postaja Ko~evje,
@aga
Rog
Novo
mesto - GGN
Ko~e Vojsko Zavi`an Plitvi~ka
jezera
Forest site classification
[umarska zajednica Ompha-
lodo-
Fagetum
Ompha-
lodo-
Fagetum
Omphalodo-
Fagetum,
Lamio orva-
lae-Arunco-
Fagetum
Ompha-
lodo-Fa-
getum
Ompha-
lodo-Fa-
getum
Ompha-
lodo-Fa-
getum
Ompha-
lodo-Fa-
getum
Country / Dr`ava SLO SLO SLO SLO SLO CRO CRO
Field methods and data analyses
Prikupljanje i analiza podataka
Three basic approaches were used in this study. First was the analysis of changes
in the total volume and beech and silver fir proportion over the research period,
which was completed for 7 locations (Table 1). The second was the examination of
the long term data on number of trees and living volume according to the diameter
at breast height (dbh) performed for 5 research sites and the third was the analysis
of regeneration in two selected old-growth forest reserves.

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
17
First information about the tree species composition in Rajhenavski rog and
Pe~ka dates back to 1883 and 1893 respectively, when the first management plans
for the area were done (Hufnagel, 1893). All the rest of the long term data were
gathered as a part of management activities after year 1950 with 10 years period
between the measurements until today. In most of the cases dbh was measured and
tree species defined for all trees in the reserve (full callipering). In the case of ^orko-
va uvala measurements were made on a systematic grid over the whole area of the
reserve two times. Also the data gathered on 1 ha plots in most representative part
of the reserve in CU and DT are included in this analysis, but just to observe the
changes in tree species composition. In all cases only trees above 10 cm in diameter
were included in the analysis. 5 cm diameter classes (up to 100 cm diameter) were
used to show the distribution of number of trees and volume per hectare according
to dbh. In cases where previous measurements were performed only up to certain
dbh, all trees above that dbh were put together in the last diameter class in all fol-
lowing data sets.
Between 2001 and 2004, under different light conditions in gaps, under stands
surrounding the gaps, and under the closed canopy a N-S oriented 5x5 m grid was
established and 773 - 1,5 x 1,5 m plots were established on the grid intersections.
On each plot all seedlings of each tree species were counted and categorized within
several height classes: 1 year-old seedlings, <20 cm height, 21 - 50 cm, 51 - 90 cm,
91 - 130 cm, 131 - 200 cm, 201 - 300 cm and >300 cm.
RESULTS
REZULTATI
Long-term changes in tree species composition and growing stock
Dugoro~ne promjene smjese vrsta drve}a i drvne zalihe
Seven old-growth forest under investigation cover substantial range from the very
start of the Dinaric mountain range in Trnovski gozd Slovenia, represented by Bukov
vrh, to the core area in Croatia, represented by ^orkova Uvala and Dev~i}a tavani. In
the 50’ the share of silver fir in growing stock of all live trees amounted on average
between 50-65% (Figure 2). Values below 30% were measured in Krokar and Bukov
vrh. First is only partly covered by beech–fir ecosystem, while Bukov vrh represents
a transition towards high mountan beech forests. However, from 1950 on, all old-
growth forests demonstrate a decrease of silver fir share in the growing stock of all
live trees. This reduction is less pronounced in Croatia compared to Slovenia. Still,
also within countries significant differences in decrease exist. The steepest decrease
took place in Pe~ka in Slovenia, where silver fir share declined from 58 % in 1893 to
17 % in 2003, which represents a 41 % decrease in about 100 years. It appears that
old-growth forest at the edge of Dinaric mountain chain (for example Pecka with 20
% decrease over the last 20 years) or at the edge of the beech-fir ecosystems within
the Dinaric mountains (Krokar with 17 % decrease in the last 40 years) demonstrate
a more noticeable decrease. The decrease of the silver fir was not constant over the

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
18
time. Two old-growth forests show a more pronounced decrease in the last interval
between the measurements, namely Pe~ka and Dev~i}a tavani, while the change is not
so evident in case of Bukov vrh and Rajhenavski Rog. The majority of others show a
decrease of ca. 5 %. Observing longer periods of time we can see different patterns,
as for example a relatively constant and lately accelerated decrease in Pecka and an
increase until 50’ followed by a gradual decrease in Rajhenavski Rog.
Figure 2: Long-term changes in proportion of silver fir in living volume in seven beech-fir old-growth
forests in Dinaric mountains
Slika 2. Dugoro~ne promjene volumnog udjela obi~ne jele u sedam dinarskih bukovo-jelovih pra{uma
The most important species competing with silver fir is beech, all the remaining
species from Norway spruce to sycamore maple represent a minor share of maxi-
mum 5 % in some reserves. The trends of beech proportion in the growing stock
of all live trees were reciprocal to those demonstrated by silver fir (Figure 3), as
expected. The beech share in all old-growth forests exceeds 40 %. The most silver
fir dominant forests remain Corkova Uvala and Rajhenavski Rog.
The growing stock of old-growth forest was in an interval form slightly above
500 m3/ha on more extreme sites as Bukov vrh, to 942 m3/ha in Pecka in 50’ (Figure
4). The average value of most representative beech-fir old-growth forests was about
800 m3/ha (RR, PE, CU), the rest of old-growth forest with lower growing-stock
include also additional less productive sites in inventoried protected compartments.
Most of the variation intervals of growing stock in the observed time are close to
100 m3/ha, which seems reasonable in regard to relative large areas under investiga-
tion (from 10 to 100 ha). The maximal decrease of growing stock was observed in
Pecka. Here it declined form 942 m3/ha in 1953 to 698 m3/ha in 2003. The highest
increase of almost 150 m3/ha over 45 years was recorded in Krokar. In some old-
growth forests (e.g. RR) the growing stock remained relatively constant over almost
60 years. In the period between last two measurements there is an noticeable trend
of growing stock decrease in all forest reserves with the exception of CU2.

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
19
Changes in tree species frequency and growing stock distributions
according to dbh
Promjene distribucija vrsta drve}a i drvne zalihe s obzirom na prsni promjer
From seven old-growth forests, five with complete data sets for the last two
inventories in 1980’ and 2000’, respectively were selected for a detailed analysis of
diameter and growing stock distributions. The diameter distributions including all
tress showed more tendency towards the rotated sigmoid than the negative expo-
Figure 3: Long-term changes in share of beech in living volume in seven beech–fir old-growth forests in
Dinaric mountains
Slika 3. Dugoro~ne promjene volumnog udjela obi~ne bukve u sedam dinarskih bukovo-jelovih
pra{uma
400
500
600
700
800
900
1000
1945 1965 1985 2005
m
3
/ha
RR
PE
KR
CU1
CU2
DT
BV
ST
Total volume
Yea r
Figure 4: Long-term changes in amount of growing stock of live trees in seven beech–fir old-growth
forests in Dinaric mountains
Slika 4. Dugoro~ne promjene prosje~ne drvne zalihe `ivih stabala u sedam dinarskih bukovo-jelovih
pra{uma

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
20
nential distribution (Figure 5). The closest to the reverse J-shaped curve was the
distribution from Corkova Uvala and Strmec, if we neglect the larger diameter sizes
above 19 dbh. The curve from ^orkova uvala was also the most stable in time, while
the curves form Slovenian old-growth forest showed changes in the shape of the cur-
ve. However, this was different from forest to forest. For example, in Pecka, Strmec
and Krokar we noticed a decrease in small diameters and an increase in large ones,
while in Rajhenavski rog there was a decrease in the middle diameters. The closer
look into the diameter distribution of each species reveals the nature of changes
for the curve of total stems in reserve. There are differences between countries and
among individual reserves. The curves for both species seem to be relatively stable
for both species in ^orkova uvala, while in Slovenia the curves for species show si-
gnificant differences in time. The most apparent and worrying trend is the decrease
of the silver fir curve along whole range of the diameter distribution. Still, this is
most pronounced in the lowest and larges diameters. On the other hand the beech
shows a decrease of curves among the youngest diameters and increase among the
largest. This trend points to the severe mortality among the young beech generation
which competed for space in the gaps created after the most severe decline of silver
fir, which was followed by two successive windthrows. The gaps were partially clo-
sed by the growth of the advanced regeneration and partially by lateral extensions
of the crowns of surrounding trees. The exception is Rajhenavski rog where the de-
crease of silver fir and increase of beech in the low diameter classes seem to be slow
and thus synchronous. In all reserves the inverse sigmoid curve was more expressed
with beech than with silver fir.
In the last two decades the growing stock curve for all live trees has shifted
towards right hand side of the figure in all Slovenian old-growth forests, whilst it
stayed relatively stable in ^orkova uvala (Figure 6). This is probably due to more
pronounced silver fir decline in Slovenia, which triggered lush ingrowth of beech.
This phenomenon is more pronounced if we analyze growing stock distributions
per dbh for beech and silver fir separately. The beech curve had the strongest shift
towards right in Pe~ka, followed by Krokar and Strmec. This curve seems relatively
stable in Rajhenavski rog and ^orkova uvala. However, the silver fir curve showed
the strongest decrease in Pe~ka, followed by Strmec and Krokar. In Rajhenavski rog
and ^orkova uvala it stayed stable with a slight move to the right hand side. To allow
comparability the diameters above dbh class 20 were merged into 20th dbh class for
Pe~ka, Rajhenavski rog and ^orkova uvala. For Krokar and Strmec all higher diame-
ters were merged into 16th dbh class. Still, from the graphs we could see the general
rule that beech and silver fir supplement each other within the curves. In the lower
diameters beech is more frequent and competitive, while silver fir compensates for
this in larger diameters (and heights). The beech curve usually finishes with 19th dbh
class, while silver fir often attains diameters up to 24th dbh class, with individual
trees up to 32th class.

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
21
Regeneration tree species composition and height distribution
Struktura pomlatka po vrstama drve}a i visinama
The density of regeneration in total was almost five times higher (62.066 per
ha) in Rajhenav compared to the ^orkova uvala (13.083 per ha) forest reserve (t =
19,4586, p = 0,0000), which is mostly due to an almost ten-fold increase (54.699
versus 5588 per ha) in the density of beech at Rajhenav (Figure 1 right). However, the
total density of silver fir was more than twice as high (6312 versus 3187 per ha) in
^orkova uvala (t = -5,7711, p = 0,0000). In spite of the fact that there was much less
silver fir in the upper storey, the density of one year old and up to 20 cm tall silver fir
seedlings in Rajhenav was higher than in ^orkova uvala. However, the density of silver
fir seedlings taller than 20 cm was higher in ^orkova uvala, as there were no seedlings
taller than 50 cm in the Rajhenav old-growth forest reserve (Figure 1 left). Lower den-
sities of beech seedlings in ^orkova uvala could be explained by lower radiation levels
0
20
40
60
3 5 7 9 11 13 15 17 19
N
beech 80
silver fi r 80
beech 03
silver fi r 03
total 80
total 03
Pecka
5 cm diameter class
0
20
40
60
3 5 7 9 11 13 15 17 19
N
beech 85
silver fi r 85
beech 07
silver fi r 07
total 85
total 07
Rajhenavski rog
5 cm diameter class
0
20
40
60
3 5 7 9 11 13 15 17 19
N
beech 87
silver fi r 87
beech 06
silver fi r 06
total 87
total 06
Corkova uvala
5 cm diameter class
0
20
40
60
3 5 7 9 11 13 15 17 19
N
beech 84
silver fir 84
beech 04
silver fir 04
total 84
total 04
Strmec
5 cm diameter class
0
20
40
60
3 5 7 9 11 13 15 17 19
N
beech 84
silver fi r 84
beech 04
silver fi r 04
total 84
total 04
Krokar
5 cm diameter class
Figure 5: Comparison of diameter distributions of five beech–fir old-growth forests according to tree
species between the last two inventories
Slika 5. Usporedba distribucija prsnih promjera za posljednje dvije inventure na primjeru pet bukovo-
jelovih pra{uma

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
22
and smaller gaps compared to Rajhenav, and also by much more intensive competition
from the herb layer in ^orkova uvala. The main reason for the dramatically low den-
sity of fir seedlings above 20 cm tall in Rajhenav is the heavy browsing pressure, due
to a high population density of red and roe deer in the Dinaric mountains in Slovenia.
This also explains the much denser coverage of ground vegetation in ^orkova uvala,
especially Rubus species, which is highly desired by deer.
DISCUSSION
RASPRAVA
Seven examined beech–fir old-growth forests covering broader geographical
scale from north-west to central Dinaric Mountains showed a significant decrease
0
20
40
60
80
100
3 5 7 9 11 13 15 17 19
beech 80
silver fi r 80
beech 03
silver fi r 03
total 80
total 03
m3/ha
Pecka
5 cm diameter class
0
20
40
60
80
100
3 5 7 9 11 13 15 17 19
beech 85
silver fi r 85
beech 07
silver fi r 07
total 85
total 07
m3/ha
Rajhenavski rog
5 cm diameter class
0
20
40
60
80
100
3 5 7 9 11 13 15 17 19
beech 87
silver fi r 87
beech 06
silver fi r 06
total 87
total 06
m3/ha
Corkova uvala
5 cm diameter class
0
20
40
60
80
100
3 5 7 9 11 13 15 17 19
beech 84
silver fir 84
beech 04
silver fir 04
total 84
total 04
Strmec
5 cm diameter class
m3/ha
0
20
40
60
80
100
3 5 7 9 11 13 15 17 19
m3/ha
beech 84
silver fi r 84
beech 04
silver fi r 04
total 84
total 04
Krokar
5 cm diameter class
Figure 6: Comparison of distributions of growing stock in relation to DBH of five beech–fir old-
growth forests between the last two inventories
Slika 6. Usporedba distribucija drvne zalihe za posljednje dvije inventure na primjeru pet bukovo-jelo-
vih pra{uma

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
23
of silver fir share in growing stock in the last half of the century. Similar trends were
reported for managed forests many times during the period of acute silver fir decli-
ne form 1950’ to mid 80’ ([afar 1964; Kandler 1992; Korpel 1985; Larsen 1986;
Leibundgut 1974; Mlin{ek 1964). The decrease of silver fir in growing stock was of
a different magnitude among individual virgin forests, still it appears to slow down
in recent decades. This is conform with the findings from managed forests were first
observations of improved health status of silver fir were reported in the last decades
(Dobrowolska 1998). The exceptions were Dev~i}a tavani and Pe~ka, the last being
hit by two successive windstorms (Nagel and Diaci 2006; Nagel et al. 2006).
In all studied forests silver fir was replaced by the beech. This is a common trend
often reported for individual reserves and managed forests (Bon~ina et al. 2003;
Diaci 1994). In spite of significant tree species replacement the overall growing
stock in the majority of old-growth forests didn’t fluctuate as expected. Exception
was the old-growth Pecka, due to above mentioned reasons.
The causes for silver fir decline are complex, however they are of natural, an-
thropogenic and mixed origin. The anthropogenic reasons include the pollution of
ecosystems on a local and global scale, climate change, high densities of the ungula-
tes and non adapted silvicultural systems to the silver fir ecology. The last, appears
not to be one of the primary causes due to the silver fir decline in old-growth forests.
However, forestry might have an effect, through building of fine forest road network
and thus changing the local forest climate, removal of CWD on which silver fir often
regenerates and changing from single tree selection to irregular shelterwood system
in Slovenia.
The most important natural cause of silver fir decline is the tree species al-
ternation in persistent (climax) communities. The word alternation of species was
proposed by a French forester in 1905 (Schaeffer and Moreau 1958) and then often
observed or researched in temperate forests (Forcier 1975; Fox 1977; Watt 1947),
including Croatia and Slovenia ([afar 1967; Ga{per{i~ 1974). Initially, the alterna-
tion was understood mostly on the local scale, from one or to at most few large ca-
nopy trees. Here, most processes affecting density and species composition operate,
except dispersal (Fox 1977). (Watt 1947) was the first wrote about this issue; the
alternation was reflected in the spatial mosaic of the community. Later (Fox 1977)
found that sapling abundance was lower beneath the canopies of the same species.
0
1000
2000
3000
4000
5000
6000
7000
oy 20 50 90 130 200 300 300-
cm
rajhenav
corkova
N/ha
0
2500
5000
7500
10000
12500
15000
oy 20 50 90 130 200 300 300-
cm
rajhenav
corkova
N/ha
Figure 7: Density (n ha-1) of fir (left) and beech (right) one year old (oy) and older seedlings in different
height classes (cm) in Rajhenav and ^orkova uvala
Slika 7. Gusto}a (n ha-1) jelovog (lijevo) i bukovog (desno) ponika i pomlatka po visinskim klasama
(cm) u pra{umama Rajhenav i ^orkova uvala

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
24
Similar results were reported also for Dinaric beech–fir forests (Ga{per{i~ 1974).
There are several mechanisms behind the alternation (Fox 1977):
1. The substrate and micro-relief variability could be important on the diverse
carstic conditions of Dinaric Alps. Short (climate, seed bed) and long-term
(nutrients, chemical properties) influences of CWD could be added to this
group, although they operate also tree species specific.
2. Demographic causes of alternation - each species inhibits the survival and
growth of its own species most severely (autoinhibition):
– parent tree is a source of predators and pathogens: herbivorous insects,
other predators, toxins
– conspecific trees my be thinned more severely: similar niche and need for
resources, e.g. soil, light with temporal variability, variability in quality
– other reasons: lover abundance of long living species - they take long to
pass a certain dangerous height range
– tolerance relations: adaptation to specific disturbance regime (small and
large gaps); expected crown diameter at death is favoring another species;
different crown size between conifers and broadleaves (silver fir - beech)
– better growth of evergreens underneath of broadleaves
– plants create individual microhabitats (light, precipitation).
However, also regional and global factors might have an important effect on the
species alternation as shown in this study. For example different densities of ungu-
lates, which hindered the regeneration of silver fir in Slovenia completely in last 30
years, and various silvicultural regimes operate on a regional scale. Moreover, the
global climate change, with higher temperatures, more severe disturbance regime is
less favorable for silver fir compared to beech. Similar replacements of both species
have been found after the last ice ages (Andric and Willis 2003; [ercelj 1996). Fi-
nally, it appears very difficult to draw a line between different scales of operation of
species alternation and natural versus anthropogenic factors.
The examined old-growth forest covered mainly large areas, however the dia-
meter distribution of individual old-growth forest did not follow the negative expo-
nential distribution as reported from many studies in old-growth forests (Lorimer
1980) (Leak 1996). The reverse negative exponential distribution suggest similar
growth and mortality rates across the diameter range. The observed curves were
closer to a rotated sigmoid distribution. This observation is not a new one (Goff and
West 1975; Hartman 1987; Leibundgut 1982; Westphal et al. 2006). The deviation
from the negative exponential curve might be due to higher mortality in lower and
higher diameter classes and/or faster diameter growth in the mean classes. It is inte-
resting that the rotated sigmoid distribution was found to provide the best fit also for
managed equilibrium state plenter beech–fir forest in Switzerland (Schütz 2001).
The curves for growing stock per dbh revealed that beech and silver fir occupy
different niche within ecosystem. In the lower diameters beech is more frequent and
competitive, while silver fir compensates for this in larger diameters (and heights).
This was already reported for two beech–fir old-growth forests in Slovenia (Hart-

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
25
man 1987; Turk et al. 1985). A general observation from most reserves was, that the
diameter and volume curves were not stable in time.
A comparison of Slovenian old-growth forests with ^orkova Uvala in Croatia
revealed a strong decrease of the silver fir curve in Slovenia along all the diameter
distribution, with peaks in the lowest and largest diameters. If we add to this also the
almost absent regeneration of silver fir in Slovenia, than it is obvious that silver fir is
significantly more endangered in Slovenia than Croatia.
The worrying trends of silver fir decline in Slovenian old-growth forests require
a continuation of special management procedures for controlling the roe and red
deer density. The different niche occupation of silver fir and beech within the same
ecosystem call for a individual silvicultural treatment of each species, e.g. different
rotation periods, target diameters, modes of regeneration. Shaded silver fir regene-
ration and mature trees of high volume and age allow silver fir to develop the full
competition potential. Still, the silver fir decline is partially also a natural phenome-
non linked to species replacement, therefore more unpredictability and flexibility
should be integrated in the silvicultural systems.
REFERENCES
LITERATURA
Andric, M., K. J. Willis, 2003: The phytogeographical regions of Slovenia: a consequence of
natural environmental variation or prehistoric human activity? Journal of Ecology. 91:
807-821.
Bon~ina, A., F. Ga{per{ic, J. Diaci, 2003: Long-term changes in tree species composition in
the Dinaric mountain forests of Slovenia. Forestry Chronicle. 79: 227-232.
Brinar, M., 1969: Vpliv svetlobe na razvoj bukovega mladja. Zbornik - Research reports. 7:
61-144.
Diaci, J., 1994: Spreminjanje naravne gozdne vegetacije ob vi{inskem gradientu Ve`e -
Dleskov{ke planote v Savinjskih Alpah. Zb. gozdarstva in lesarstva: 45-84.
Dobrowolska, D., 1998. Structure of silver fir (Abies alba Mill.) natural regeneration in the
’Jata’ reserve in Poland. Forest Ecology and Management. 110: 237-247.
Ficko, A., A. Boncina, 2006: Silver fir (Abies alba mill.) distribution in Slovenian forests. Re-
search Reports - Forestry and Wood Science and Technology, 79: 19-35.
Forcier, L. K., 1975: Reproductive Strategies and Co-Occurrence of Climax Tree Species.
Science. 189: 808-810.
Fox, J. F., 1977: Alternation and coexistence of tree species. The American Naturalist. 111:
69-89.
Ga{per{i~, F., 1974: Zakonitosti naravnega pomlajevanja jelovo-bukovih gozdov na visokem
krasu Sne`ni{ko-Javorni{kega masiva. In Strokovna in znanstvena dela. Biotehni{ka
fakulteta, Ljubljana. p. 133.
Goff, F. G., D. West, 1975: Canopy-understorey interaction effects on forest population
structure. For. Sci. 21: 98–108.
Hartman, T., 1987: Gozdni rezervati Slovenije - Pragozd Rajhenavski Rog. Univerza Edvarda
Kardelja v Ljubljani, Biotehni{ka fakulteta, Ljubljana, p. 99.
Hufnagel, L., 1893: Wirtschaftsplan der Betriebsklasse III. Hornwald. Ko~evje
Kandler, O., 1992: No relationship between fir decline and air pollution in the Bavarian For-
est. For. Sci. 38: 866-869.

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
26
Kordi{, F., 1985: Gozdni rezervati Slovenije - Pragozd Bukov vrh, Univerza Edvarda Kardelja
v Ljubljani, Biotehni{ka fakulteta, Ljubljana, 77 pp.
Korpel, S., 1985: Stav a vyvoj jedle na Slovensku vo vztahu k jej odumieraniu. Acta Fac. For.
27: 79-102.
Larsen, B. J., 1986: Das Tannensterben: Eine neue Hypothese zur Klärung des Hintergrundes
dieser Rätselhaften Komplexkrankheit der Weisstanne (Abies alba Mill.). Forstw. Cbl.
105: 381-396.
Leak, W. A., 1996: Long-term structural change in uneven-aged northern hardwoods. For.
Sci. 42: 160–165.
Leibundgut, H., 1974: Zum Problem des Tannensterbens. Schweizerische Zeitschrift für
Forstwesen. 125: 476 - 484.
Leibundgut, H., 1982: Europäische Urwälder der Bergstufe. Haupt, Bern. p. 308.
Lorimer, C. G., 1980: Age Structure and Disturbance History of a Southern Appalachian
Virgin Forest. Ecology. 61: 1169-1184.
Mlin{ek, D., 1964: Su{enje jelke v Sloveniji - prvi izsledki. Gozdarski vestnik. 22: 145-159.
Mlin{ek, D., 1967: Pomlajevanje in nekatere razvojne zna~ilnosti bukovega in jelovega mladja
v pragozdu na Rogu=Verjüngung und Entwicklung der Dickungen im Tannen-Buchen
Urwald “Rog” (Slowenien). Zbornik BF. 15: 7-32.
Nagel, T. A., J. Diaci, 2006: Intermediate wind disturbance in an old-growth beech-fir forest
in southeastern Slovenia. Canadian Journal of Forest Research-Revue Canadienne De
Recherche Forestiere. 36: 629-638.
Nagel, T. A., M. Svoboda, J. Diaci, 2006: Regeneration patterns after intermediate wind dis-
turbance in an old-growth Fagus-Abies forest in southeastern Slovenia. Forest Ecology
and Management. 226: 268-278.
Ro`enbergar, D., 2000: Razvojne zna~ilnosti sestojev v pragozdnih ostankih Pe~ka in Ra-
jhenavski Rog = Stand dynamics of the virgin forest remnants of Pe~ka and Rajhenavski
Rog. Gozdarski vestnik 58: 53-55.
Schaeffer, R., R. Moreau, 1958: L’alternance des essences. In Soc. Forest. France-Compté
Bull. 1-12, 76-84, pp. 277-298.
Schütz, J. P., 2001: Der Plenterwald und weitere Formen strukturierter und gemischter Wäl-
der. Parey, Berlin, p. 207.
[afar, J., 1964: Problem su{enja jele i na~ina gospodarenja na Macelj Gori. [umarski list: 1-
16.
[afar, J., 1967: Izmjena bukve i jele na panonskim gorama. [umarski list: 291-298.
[ercelj, A., 1996: The origins and development of forests in Slovenia. In Opera. Academia
scientiarum et artium Slovenica, Classis IV: Historia naturalis, Ljubljana.
Turk, V., A. Kastelic, T. Hartman, 1985: Gozdni rezervati Slovenije - Pragozd Pe~ka. Univerza
Edvarda Kardelja v Ljubljani, Biotehni{ka fakulteta, Ljubljana, p. 75.
Watt, A. S., 1947: Pattern and process in the plant community.
Westphal, C., N. Tremer, G. V. Oheimb, J. Hansen, K. V. Gadow, W. Hardtle, 2006: Is the
reverse J-shaped diameter distribution universally applicable in European virgin beech
forests? Forest Ecology and Management. 223: 75-83.
Zeibig, A., 2001: Lückenstörungmuster eines montanen illyrischen Buchenurwaldes der As-
soziation Isopyro-Fagetum im Reservat Krokar der Borov{ka Gora im südwestlichen Slo-
wenien - Diplomarbeit. Dresden, Fakultät Forst-, Geo- und Hydrowissenschaften der
Technischen Universität Dresden, str. 92.

J. Diaci, D. Ro`enbergar, S. Mikac, I. Ani}, T. Hartman, A. Bon~ina: Long-term changes in tree species
composition in old-growth dinaric beech-fir forest. Glas. {um. pokuse, Vol. 42, 13 – 27, Zagreb, 2007–08.
27
DUGORO^NE PROMJENE U SASTAVU VRSTA DRVE]A
DINARSKIH BUKOVO-JELOVIH PRA[UMA
SA@ETAK
Uzastopne inventure dinarskih bukovo-jelovih {uma tijekom pro{log stolje}a su
pokazale promjene u njihovoj strukturi. Te su spoznaje va`ne za razumijevanje pro-
cesa pomla|ivanja u pra{umama i gospodarenje {umama. Ipak, nijedna studija nije
uspore|ivala nizove podataka iz nekoliko pra{uma. Analizirali smo podatke o drvnoj
zalihi iz 31 inventure u sedam pra{uma koje se prostiru od sjeverozapadnih Dinarida
u Sloveniji do njihova sredi{njeg dijela u Hrvatskoj. U svim je {umama ustanovlje-
no smanjenje udjela obi~ne jele u drvnoj zalihi. Smanjenje udjela jele je razli~ito
me|u pojedinim pra{umama. Ukupna drvna zaliha u ve}ini pra{uma ne pokazuje
o~ekivane promjene s obzirom na odumiranje jele. Distribucija prsnih promjera u
pet odabranih pra{uma bli`a je obrnuto sigmoidalnoj nego negativno eksponencijal-
noj funkciji, unato~ relativno velikom podru~ju uzorkovanja. Distribucije volumena
po debljinskim stupnjevima pokazuju kako bukva i jela zauzimaju razli~ite ni{e unu-
tar ekosustava. Bukva je vi{e zastupljenija u ni`im debljinskim stupnjevima, dok jela
kompenzira taj prostor u vi{im debljinskim stupnjevima i visinama. Obje krivulje za
sve istra`ivane rezervate pokazuju nestabilnost u vremenu. Najjasniji i zabrinjavaju}i
trend u Sloveniji je opadanje udjela jele unutar ~itavog opsega distribucije prsnih
promjera, s najvi{om to~kom u ni`im i vi{im stupnjevima. Osim toga, pomlatka jele
iznad 20 cm visine nema zbog visoke gusto}e populacije divlja~i. ^orkova uvala u
odnosu na sve ostale istra`ivane pra{ume ~ini se najuravnote`enijom s obzirom na
analizirane parametre. Rezultati pokazuju generalni istovremeni trend zamjene jele
obi~nom bukvom na ve}oj geografskoj razdiobi. U radu su prikazani mogu}i razlozi
tih promjena te su date preporuke za uzgajanje bukovo-jelovih {uma.
Klju~ne rije~i: pra{uma, bukovo-jelove {uma, distribucija prsnih promjera, drvna
zaliha, sastojinski oblik, izmjena vrsta drve}a, dugoro~ne izmje-
re.