Content uploaded by Miroslav Svoboda
Author content
All content in this area was uploaded by Miroslav Svoboda
Content may be subject to copyright.
Gap regeneration and replacement patterns
in an old-growth Fagus–Abies forest
of Bosnia–Herzegovina
Thomas A. Nagel •Miroslav Svoboda •
Tihomir Rugani •Jurij Diaci
Received: 31 July 2009 / Accepted: 25 November 2009 / Published online: 11 December 2009
ÓSpringer Science+Business Media B.V. 2009
Abstract We examined the influence of small-scale
gap disturbances on stand development and tree
species coexistence in an old-growth Fagus sylvati-
ca–Abies alba forest in the Dinaric Mountains of
Bosnia–Herzegovina. The structure and composition
of tree regeneration in gaps were compared to the
forest as a whole, and the influence of gap size on the
density and composition of regeneration was
assessed. Transition probabilities were also calcu-
lated from gapfillers in different life stages to
examine canopy replacement patterns. The structure
and composition of tree regeneration were similar
between gaps and the forest as a whole, and there was
no relationship between overall regeneration density
and gap size, indicating most individuals established
prior to gap formation. Likewise, there was no strong
evidence of gap-size partitioning for shade tolerant
F. sylvatica and A. alba, although less tolerant
Acer pseudoplatanus only recruited to taller life
stages in larger gaps. Transition probabilities calcu-
lated from the seedling and sapling data suggest that
most gaps will be captured by F. sylvatica, while
probabilities based on pole-sized gapmakers indicate
both A. alba and F. sylvatica will be maintained in
the canopy. We suggest that gaps primarily play a
role in reorganizing advance regeneration, and that
coexistence of shade tolerant F. sylvatica and A. alba
is more likely related to their differential ability to
tolerate shaded understory conditions, particularly
during larger life stages, rather than gap-size
partitioning.
Keywords Abies alba Canopy gap
Disturbance Fagus sylvatica Gap size
Species coexistence
Introduction
Numerous studies have examined the importance of
canopy gaps for maintaining tree diversity (de Ro
¨mer
et al. 2007; Gutierrez et al. 2008; Lertzman 1992;
Rebertus and Veblen 1993; Taylor 1990; Uhl et al.
1988; Veblen 1986). Disturbances that open small
canopy gaps create environmental heterogeneity,
particularly the amount of light penetrating to the
forest understory, and thereby provide a range of
regeneration niches for species with different life
history attributes. In particular, tree species with
different levels of shade tolerance may differ in their
ability to exploit gaps of various sizes (Brokaw 1985;
Denslow 1980; Whitmore 1989). For example, many
T. A. Nagel (&)T. Rugani J. Diaci
Department of Forestry and Renewable Forest Resources,
Biotechnical Faculty, University of Ljubljana,
Vecna Pot 83, 1000 Ljubljana, Slovenia
e-mail: tom.nagel@bf.uni-lj.si
M. Svoboda
Faculty of Forestry and Wood Sciences, Czech University
of Life Sciences Prague, Kamy
´cka
´129, 16521 Prague
6-Suchdol, Czech Republic
123
Plant Ecol (2010) 208:307–318
DOI 10.1007/s11258-009-9707-z
studies have found that large gaps are important for
the maintenance of shade intolerant species (e.g.,
Barden 1981; Busing and White 1997; Kneeshaw and
Prevost 2007; Runkle 1982).
It is less clear how gaps influence the coexistence
of shade tolerant species. Some authors have argued
that partitioning of gaps by size plays a minor role in
the maintenance of tolerant species, and suggest that
diversity is simply a function of higher densities in
gaps compared to the shaded forest understory
(Busing and White 1997; Denslow 1995). However,
because gap-filling trees thin as they compete for
space in the canopy, diversity in gaps may only be a
temporary outcome of increased density (Hubbell
et al. 1999; Schnitzer and Carson 2001). Furthermore,
when a sufficient amount of light is present, shade
tolerant species often establish a bank of seedlings
and saplings under the forest canopy, such that gap
formation may mainly play a role in reorganizing
advance regeneration through subsequent recruitment
to larger life stages (Lertzman 1992; Nagel et al.
2006; Uhl et al. 1988; Webb and Scanga 2001). Thus,
the influence of canopy disturbance on forest dynam-
ics largely depends on the differential ability of tree
species to tolerate shaded understory conditions
(Canham 1989).
In the mountains of central and southeastern
Europe, forests are often dominated by mixtures of
Fagus sylvatica L. and Abies alba Mill. The forma-
tion of small to intermediate scale (i.e., 0.005–0.5 ha)
gaps is the dominant natural disturbance process
throughout the region (Dro
¨sser and von Lu
¨pke 2005;
Nagel and Diaci 2006; Nagel and Svoboda 2008).
However, few studies have examined the influence of
gap disturbances on patterns of tree regeneration
(Nagel et al. 2006; Rozenbergar et al. 2007). Such
studies may provide important insight into the
dynamics of these forests, as well as the mechanisms
that promote tree species coexistence.
Understanding the coexistence of F. sylvatica and
A. alba is rather challenging given that both species
have similar life history attributes. During early life
stages, both are highly shade tolerant (Ellenberg
1988), and seedlings of both species establish on a
variety of microsite conditions (Szewczyk and
Szwagrzyk 1996). Thus, it is unclear whether inter-
specific differences in regeneration niches (e.g.,
Grubb 1977), particularly niche differentiation along
a gap-size gradient, promote coexistence of
F. sylvatica and A. alba. Furthermore, the scant age
data from old-growth forests do not indicate a
substantial difference in the life span of both species
(Firm et al. 2009; Mlinsek 1967; Nagel et al. 2007;
Piovesan et al. 2005), so that an interspecific
difference in longevity is an unlikely cause of
coexistence (e.g., Lertzman 1995; Lusk and Smith
1998). A potential coexistence mechanism could be
related to ontogenetic changes in shade tolerance
(Kneeshaw et al. 2006; Lusk 2004). Interspecific
differences in shade tolerance may be more pro-
nounced at the pole-sized tree stage, because the ratio
of photosynthetic to non-photosynthetic biomass
decreases with increasing plant size (Givnish 1988).
Therefore, it is likely that pole-sized trees are only
able to recruit to larger stages if sufficient light is
available, which is not only dependent on gap size,
but also the spatial and temporal pattern of gap
formation (Canham 1989).
In this study, we examine how canopy gap
disturbances influence forest dynamics and tree
species coexistence in an old-growth Fagus–Abies
forest in the Dinaric Mountains, Bosnia and Herz-
egovina. Specifically, our objectives were to compare
the structure and composition of regeneration across a
range of tree-fall gap sizes and the forest as a whole;
and 2) to identify species-by-species replacement
patterns in gaps across different tree life stages.
Study area
We studied old-growth F. sylvatica–A. alba stands in
the Perucica Forest Reserve in the Dinaric Moun-
tains, Bosnia–Herzegovina. The reserve is approxi-
mately 1,400 ha, making it one of the largest areas of
old-growth in southeastern Europe, and thus affords a
unique opportunity to study natural forest structure
and dynamics. The climate has both the Mediterra-
nean and continental influence, with a mean monthly
temperature range of -3.3°C in January to 15.0°Cin
July and a mean annual precipitation of 1,837 mm
(Cemerno station—1,305 m). Terrain in the reserve is
characterized by deeply dissected, moderate to steep
slopes that descend into the Perucica river. The main
parent materials include limestone and dolomite on
the slopes and cliffs bordering the reserve and a
mixture of acidic sandstone and shale at lower
elevations. Therefore, soils are diverse on some sites,
308 Plant Ecol (2010) 208:307–318
123
particularly where calcareous soils have eroded down
slopes (Fukarek and Stefanovic 1958).
Between approximately 1,000 and 1,600 m in
elevation, F. sylvatica and A. alba are the main
canopy species (Fukarek and Stefanovic 1958).
Subdominant canopy species include maple
(Acer pseudoplatanus L.), elm (Ulmus glabra Huds.),
ash (Fraxinus excelsior L.), and Norway spruce
(Picea abies (L.) Karst). We sampled four stands in
the mixed F. sylvatica–A. alba vegetation zone
between 1,000 and 1,300 m in elevation. Three of
the stands (Tunjemir 1, 2, and 3) varied in aspect from
north to northeast and were on slopes between 20°–
30°, while the fourth stand (Zanoglina) was southwest
in aspect on a more moderate slope (16°). Formation
of small- to intermediate-sized gaps is the dominant
disturbance process in the study area; there was no
evidence of high-severity, stand replacing disturbance
events in the recent past (Nagel and Svoboda 2008).
With the exception of the basic structural character-
istics, the analyses of regeneration and canopy
replacement patterns reported in this article are based
on the combined data set from all four stands.
Methods
We established three transects in each stand varying
in length from 200 to 400 m. Transects followed
constant elevational contours and were separated by
50 m to avoid duplicate sampling. Forest structure
was sampled in 400 m
2
circular plots (n=55) placed
at 40 m intervals along each transect. In each plot, we
measured the diameter at breast height (dbh) of all
live and dead standing trees C5 cm dbh and counted
seedlings (0.5–1.3 m tall) and saplings ([1.3 m tall
and \5 cm dbh) of each species. Trees were addi-
tionally placed into a pole-sized (\20 m tall) or
canopy ([20 m tall) height class.
The same transects described above were used to
sample canopy gaps (n=87). We defined gaps as
openings in the forest canopy [5m
2
caused by the
mortality of a tree [25 cm dbh. Mortality of smaller
stems was not considered large enough to create a
canopy opening. Certainly, canopy trees [25 cm dbh
are likely to create openings larger than 5 m
2
, but old
gaps are likely to have decreased in size due to lateral
closure from surrounding canopy trees, so we chose a
5-m
2
cut-off to avoid missing these older gaps. A gap
was considered closed when stems within the gap
reached a height of 20 m (equivalent to a dbh of ca.
25 cm), equal to approximately half the height of the
surrounding canopy. As we were concerned with the
influence of gaps on tree regeneration, we focused on
the expanded gap (sensu Runkle 1982), defined as the
canopy gap plus the area delineated by the boles of
the canopy trees surrounding the gap. The expanded
gap is useful, because it encompasses the area that
both directly and indirectly influences understory
vegetation. Each expanded gap (inclusive of the
canopy gap) intersected by a transect was sampled.
We precisely measured the size of each expanded gap
by measuring radii (distance and direction) from the
approximate gap center to the bole of each tree that
defined the gap. In three of the stands, we encoun-
tered large ([0.15 ha), ‘‘messy’’ canopy openings
(n=5) characterized by several interconnected gaps
with scattered canopy trees in the gap interiors. The
boundary of these areas was usually easy to differ-
entiate from the surrounding intact canopy. The sizes
of these areas were calculated by fitting their longest
lengths and longest perpendicular widths to the
formula for the area of an ellipse.
In each expanded gap, we identified the species of
each gapmaker(s). In order to assess regeneration
(gapfillers), we measured the dbh of all pole-sized
trees and recorded the number of seedlings and
saplings by species in each expanded gap (except for
the five largest gaps [0.15 ha). Additionally, we
selected the gapfiller most likely to replace each
gapmaker. These definitive gapfillers (sensu Lertz-
man 1992) were chosen based upon their height,
location, and a visual assessment of their overall
condition. In most cases, we could easily designate
the definitive gapfillers because they were consider-
ably taller than other gapfillers, and they often
showed a height growth release in response to gap
formation that was easily seen in the field, especially
for A. alba. Most definitive gapfillers in this study
were pole-sized trees 10–20 m tall. In recently
formed gaps, however, we occasionally selected a
large sapling as a definitive gapfiller, but only if it
was substantially taller than the other gapfillers. In
many of the small or recently formed gaps, there were
no gapfillers or only seedlings and saplings. For the
small gaps (i.e., \50 m
2
) lacking gapfillers, it was
obvious that lateral crown growth of border trees
would close these openings barring subsequent gap
Plant Ecol (2010) 208:307–318 309
123
expansion. However, in several of the larger, multiple
tree-fall gaps, some of the gapmakers occasionally
lacked a definitive gapfiller.
Analysis
We tested for differences in the stem density of each
species at different life stages (i.e., seedlings, sap-
lings, and pole trees) between expanded gaps and
stand structure plots using Mann–Whitney Utests.
Since the stand structure plots were not explicitly
placed under closed canopy areas, but rather at
systematic locations along the transects, they repre-
sent the overall forest conditions, including gaps,
closed canopy, and areas intermediate between these
canopy states. This way, we were able to compare the
influence of gaps versus the forest as a whole on
regeneration patterns, rather than only testing extreme
differences between gap and non-gap areas (Battles
and Fahey 2000; Lieberman et al. 1989).
The relationship between expanded gap size and
regeneration density was examined with Spearman’s
qcorrelations. Correlations were performed sepa-
rately for the dominant species in the regeneration
layer and for total regeneration density at different
life stages (i.e., seedlings, saplings, and pole trees).
Additionally, we determined whether the species
composition of the regeneration varied with expanded
gap size. In order to do this, we examined the size
distribution of expanded gaps that contained at least
one individual of a particular species. In our case, we
focused on the two dominant canopy species, A. alba
and F. sylvatica, as well as A. pseudoplatanus, which
was the most abundant of the less shade tolerant
species present and was expected to exploit larger
gaps.
In order to examine patterns of canopy replace-
ment, we constructed four matrices of transition
probabilities based on the proportions of gapfillers
to gapmakers in each expanded gap. The first three
matrices were constructed from the relative frequency
of seedlings, saplings, and pole trees, respectively,
and the fourth matrix was based on the definitive
gapmaker-gapfiller pairs. We calculated more than
one matrix because canopy replacement patterns
based on different life stages may provide insight
into species-specific survivorship and coexistence
(White et al. 1985). The four matrices were calculated
for all the expanded gaps in study, as well as for three
expanded gap-size classes (i.e., small: \250 m
2
;
medium: 250–400 m
2
; and large: [400 m
2
).
Results
All stands were dominated by F. sylvatica and
A. alba, although the density and basal area of each
species varied among them (Table 1). In the three
Tunjemir stands, F. sylvatica was more abundant
than A. alba in the canopy layer, while A. alba
dominated the canopy in the Zanoglina stand (Fig. 1),
where it made up 79% of the total basal area. Over all
the stands, however, small F. sylvatica trees
([15 cm dbh) were several fold more abundant than
A. alba, which is clearly reflected in the diameter
distributions (Fig. 1). Neither of the two dominant
species exhibited a reverse J shape diameter distri-
bution. A. alba trees were most abundant in diameter
classes between 15–35 cm, except for at tunjemir 2,
while the distributions of F. sylvatica showed more
irregular patterns for size classes [15 cm. F. sylvat-
ica was the most abundant species in the regeneration
layer (seedlings and saplings) over all the stands and
ranged from 714 to 1,841 stem ha
-1
(Table 1).
A. alba regeneration was sparse, ranging from 5 to
155 stem ha
-1
.A. pseudoplatanus regeneration was
also present in all the four stands, with densities
similar to or greater than A. alba.
Similar to the forest as a whole, F. sylvatica was
the most abundant gapfiller across all life stages,
although A. alba gapfillers were also abundant in the
pole tree stage (Table 2). A. pseudoplatanus was the
second most abundant gapfiller species in the
seedling and sapling stages, but was considerably
less abundant in the pole tree stage. The structure and
composition of regeneration were similar between the
forest as a whole and expanded gap areas (Table 2).
None of the six species present in the regeneration
layer had significantly different densities of seed-
lings, saplings, or pole-sized trees between the two
areas (Mann–Whitney Utests; P[0.05 for each).
Furthermore, the dominant species in the regenera-
tion layer occurred in the forest as a whole and
expanded gaps at similar frequencies over the differ-
ent life stages.
Total stem density and expanded gap size were not
significantly related for the seedling (Spearman’s
310 Plant Ecol (2010) 208:307–318
123
q=0.013, P=0.909), sapling (Spearman’s q=
0.000, P=0.999), and pole tree (Spearman’s
q=0.162, P=0.147) life stages (Fig. 2). When
the relationship between density and expanded gap
size was tested for individual species at different life
stages, several of the cases were significant. The
density of A. alba seedlings increased with expanded
gap size (Spearman’s q=0.238, P=0.031),
although this data set was dominated by zero values
because seedlings occurred in only 12 gaps. There
were no consistent relationships between density and
expanded gap size for saplings and pole-sized trees of
A. alba. For F. sylvatica, the densities of seedlings
and saplings were uncorrelated with expanded gap
size, while the density of pole-sized trees showed a
positive relationship (Spearman’s q=0.376, P=
0.001). Finally, the densities of A. pseudoplatanus
seedlings, saplings, and pole-sized trees, respectively,
were found to increase with expanded gap size
(Spearman’s q=0.420, 0.316, and 0.292 respec-
tively, all P\0.01), but the sapling and pole tree
data sets were again zero-inflated.
In order to further inspect if species exploit gaps of
different sizes, we examined the size distribution of
expanded gaps that contained at least one stem of
F. sylvatica,A. abies, and A. pseudoplatanus across
different life stages (Fig. 3). In all the cases, the size
distributions of expanded gaps containing F. sylvat-
ica and A. alba were similar, indicating that the two
species do not exploit gaps of different size.
Table 1 Forest characteristics of the four study stands
Stand Trees (N ha
-1
) Basal area (m
2
ha
-1
) Saplings (N ha
-1
) Seedlings (N ha
-1
)
Live Standing dead Live Standing dead
Tunjemir 1
Fagus sylvatica 290 19 33 3 326 1,030
Abies alba 123 22 11 2 5
Acer pseudoplatanus 3\12263
Fraxinus excelsior 33
Total 419 41 47 6 348 1,098
Tunjemir 2
Fagus sylvatica 272 20 33 1 488 226
Abies alba 92 10 16 5 62 72
Picea abies 3\113
Acer pseudoplatanus 10 \11049
Ulmus glabra 3\1
Fraxinus excelsior 7
Total 377 33 49 6 560 367
Tunjemir 3
Fagus sylvatica 257 19 35 4 1,160 681
Abies alba 210 34 16 3 19 136
Picea abies 22
Acer pseudoplatanus 49 90
Total 467 53 51 7 1,228 928
Zanoglina
Fagus sylvatica 310 6 16 1 729 987
Abies alba 277 48 63 9 6 1
Picea abies 11 1 1 \14 9
Acer pseudoplatanus 1\1440
Ulmus glabra 1\11
Total 601 55 80 10 744 1,037
Plant Ecol (2010) 208:307–318 311
123
However, there was evidence that A. pseudoplatanus
recruited to larger life stages only in larger gaps. For
example, saplings only occurred in expanded gaps
[270 m
2
and pole-sized trees and definitive gapfil-
lers in expanded gaps [440 m
2
.
We identified 322 gapmakers in the 87 gaps
encountered in the study. F. sylvatica (n=128) and
A. alba (n=183) comprised most of the gapmakers,
whereas A. pseudoplatanus (n=4), P. abies (n=3),
U. glabra (n=3), and F. excelsior (n=1) gapmak-
ers were infrequent (Nagel and Svoboda 2008). The
last three species were also rarely encountered as
gapfillers and were not included in the results of
canopy replacement patterns presented here. For the
combined gap data set, transition probabilities varied
across the life stage and definitive gapfiller categories
(Table 3). At the seedling and sapling stages, there
was a high probability of F. sylvatica gap capture for
the dominant gapmaker species (ranging from 75 to
94%), whereas the probability of replacement by
A. alba was very low (ranging from 2 to 7%). There
were also low probabilities of replacement by
A. pseudoplatanus at this stage, although gapmakers
were generally more likely to be replaced by Acer than
Abies. At the pole tree stage, replacement by A. alba
was similar to that of F. sylvatica, although there was
a stronger probability of reciprocal than self-replace-
ment for both species. Similarly, the analysis of
definitive gapfillers also showed a higher probability
of reciprocal replacement between F. sylvatica and
A. alba than self-replacement, although there was a 28
and 36% probability that Fagus and Abies gapmakers,
respectively, would not be replaced by any gapfillers.
Finally, Fagus and Abies were replaced by A. pseudo-
platanus with low probabilities (ranging from 1 to 4%)
at the pole tree and definitive gapfiller stages.
The calculation of transition probabilities within
different expanded gap-size classes showed similar
0
10
20
30
40
0
20
40
60
80
100
120
140
Abies alba Fagus sylvatica
(b) tunjemir 2
0
20
180
200
0
10
20
30
40
50
(a) tunjemir 1
Abies alba Fagus sylvatica
10 30 50 70 90 110
0
20
40
60
80
100
10 30 50 70 90 110 10 30 50 70 90 110
10 30 50 70 90 110 10 30 50 70 90 110
10 30 50 70 90 110
0
20
40
60
80
100
120
(c) tunjemir 3
Abies alba Fagus sylvatica
10 30 50 70 90 110
0
20
240
260
10 30 50 70 90 110 130
0
10
20
30
40
50
60 Abies alba Fagus sylvatica
(d) zanoglina
dbh class midpoint (cm)
Number of trees / ha
Live Dead
Fig. 1 Size distributions of living and dead Abies alba and
Fagus sylvatica trees ([5 cm dbh) in the four study stands
Table 2 Density and frequency of seedlings, saplings, and pole-sized trees in the forest as a whole (For) and gaps
Species Seedlings Saplings Pole trees
Density (N ha
-1
) Frequency (%) Density (N ha
-1
) Frequency (%) Density (N ha
-1
) Frequency (%)
For Gap For Gap For Gap For Gap For Gap For Gap
Fagus sylvatica 791 1,104 75 82 667 952 84 83 207 188 80 72
Abies alba 43 35 11 15 18 7 16 11 99 123 75 85
Acer pseudoplatanus 57 76 40 39 18 16 15 17 3 2 4 6
Picea abies 10 8 16 11 1 1 7 4 4 5 11 7
Ulmus glabra 0\10 2 \1\121 0\102
Fraxinus excelsior 14290\101 0 000
Total 902 1,228 80 88 705 978 85 85 312 319 98 98
312 Plant Ecol (2010) 208:307–318
123
patterns to the overall data set, particularly for the
seedling and sapling stages (Table 3). However, there
were notable differences for the pole tree and
definitive gapfiller categories among the three
expanded gap-size classes. In general, F. sylvatica
gap capture increased in openings [400 m
2
, while
gap capture by A. alba was highest in the small- and
medium-sized openings. There were also a larger
number of gapmakers without gapfillers in the small-
and medium-sized expanded gaps, which may
indicate that lateral crown expansion is an important
gap closure process in these openings.
Discussion
The similar regeneration density observed between
gaps and the forest as a whole shows that gaps are not
necessarily primary sites of regeneration as some
studies suggest (Busing and White 1997; Denslow
1995). Our results indicate that the gap-filling process
in Perucica is mainly controlled by advance regen-
eration rather than post-treefall establishment of
seedlings. The predominance of shade tolerant
advance regeneration found in Perucica is common
to other F. sylvatica–A. alba forests in central and
southestern Europe (Nagel et al. 2006; Szwagrzyk
and Szewczyk 2008), as well as numerous temperate
and tropical forest types regulated by small-scale
canopy gaps (Lertzman 1992; Runkle 1982; Stewart
et al. 1991; Uhl et al. 1988). We also found that the
density of the dominant species remained similar
between gaps and the forest as a whole across the
seedling, sapling, and pole tree life stages, although
we expected higher densities at larger life stages in
gaps because of the prediction that growth into larger
life stages requires more light (Givnish 1988). It
seems that the high gap fraction in Perucica (14%,
Nagel and Svoboda 2008) provides sufficient light for
survival of saplings and pole-sized trees in the forest
as a whole. The stratum of advance regeneration also
explains the lack of a relationship between regener-
ation density and gap size. It is worth noting,
however, that gap age may also influence regenera-
tion density, particularly because gapfillers are likely
to thin as gaps age, but we lack gap age data to
examine this in more detail.
A potential explanation of tree species coexistence
in gap-phase forests is based on niche partitioning
along a gap-size gradient, whereby shade-intolerant
species are more competitive in larger gaps (Denslow
1980; Whitmore 1989), yet it is less likely that gap-
size partitioning occurs with shade tolerant species
(Busing and White 1997; Canham 1989). In this
study, gap size had little influence on the occurrence
of shade tolerant F. sylvatica and A. alba because
they were already established prior to gap formation.
Seedlings, saplings, and pole-sized trees of both
species occurred across the entire size range of gaps
X
XX
X
X
X
X
XX
X
XX
XX
X
X
XX
X
X
XX
XXXX
XXXX
XX
X
XX
X
X
XX
XX
XX
XX
XX
X
X
XX
X
X
X
X
X
X
X
X
X
X
X
X
X
X
XX
X
X
XXXX
X
X
X
XX
XX
X
X
0
0. 2
0. 4
0. 6
0. 8
1
1. 2
1. 4
1. 6
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
XXX
X
X
X
X
X
X
XX
X
X
X
X
X
X
X
X
X
X
X
X
X
XX
X
X
X
X
X
X
X
X
X
X
X
X
XX
X
X
XX
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
XX
X
X
0
0.02
0.04
0.06
0.08
0.1
0.12
0 200 400 600 800 1000 1200 1400 1600
X
X
X
X
XX
X
XX
X
X
X
XX
X
XX
X
X
X
XXXX XX
X
XX
X
X
X
X
XX
X
X
X
X
XX
X
X
XX
X
X
X
X
XX
X
X
X
X
X
X
X
X
X
X
XX
X
X
XX
X
X
XXX
X
X
X
XXX
XX
X
X
0
0.2
0.4
0.6
0.8
1
Density (N/m2)
Expanded
g
ap size (m2)
seedlings
saplings
pole trees
Fig. 2 Density of tree seedlings, saplings, and pole-sized
gapfillers growing in expanded gaps of various size
Plant Ecol (2010) 208:307–318 313
123
encountered in the study. However, we did find that
the density of pole-sized F. sylvatica gapfillers
increased with gap size, indicating that recruitment
to larger life stages is more successful in larger gaps.
Of the dominant species, only A. pseudoplatanus
showed compelling evidence of gap-size partitioning.
The density of A. pseudoplatanus regeneration
increased with gap size for all three life stages, and
recruitment into pole-sized trees only occurred in
gaps [440 m
2
. These findings are consistent with the
life history characteristics of A. pseudoplatanus,
which is known to be less shade tolerant than
F. sylvatica and A. alba (Ellenberg 1988; Petritan
et al. 2007), and more light demanding as regener-
ation reaches larger life stages (Hein et al. 2009).
Replacement patterns based on the seedling and
sapling data clearly reflect the dominance of F. sylv-
atica in these regeneration stages. Observations based
solely on these early life history stages would suggest
increased dominance of F. sylvatica in the future.
However, replacement probabilities based on the pole
tree and definitive gapfiller data suggest that A. alba
will be maintained in the canopy layer. Presumably,
predictions of future canopy composition based on
the identification of gap successor trees are more
realistic than predictions based on the structure of
seedlings and saplings growing in gaps (Busing 1996;
White et al. 1985). It should be noted, however, that
many factors can potentially affect the probability of
a currently observed gapfiller reaching a canopy
position. For example, natural disturbances and
climate changes may influence future replacement
probabilities, so that any predictions of future canopy
composition should only be considered realistic under
the assumption of a stable environment (White et al.
1985).
1600
1200
800
400
0
seedlings 1600
1200
800
400
0
saplings
Ac psAb alFa sy
1600
1200
800
400
0
pole trees 6000
5000
4000
3000
2000
1000
0
definitive gapfillers
Ac psAb alFa sy
Species
Expanded gap size (m2)
67
12 32
68 914
59 70
5
31 57
8
Fig. 3 Boxplots of the size distribution of expanded gaps
containing at least one sapling, pole-sized tree, and definitive
gapfiller of Fagus sylvatica (Fa sy), Abies alba (Ab al), and
Acer pseudoplatanus (Ac ps). The box represents the inter-
quartile range, and the horizontal line inside the box shows the
median. The whiskers extend to the lowest and highest values
below and above the first and third quartile, respectively,
excluding outliers. Circles and asterisks represent outliers that
are more than 1.5 and 3 times the interquartile range,
respectively. Numbers beside the boxes show the sample size.
Note that the definitive gapfiller plot has a different scale
because it includes the five large gaps where only definitive
gapfillers were recorded
314 Plant Ecol (2010) 208:307–318
123
These results also provide insight into the different
life history strategies of the studied tree species. For
example, the increased gap capture by A. alba at
larger life stages may reflect its ability to survive
longer periods of suppression than F. sylvatica,
resulting in a pool of pole-sized A. alba to capture
gaps after they form. An alternative explanation is
that the present pool of pole-sized A. alba in the
study area is a legacy of past conditions when
regeneration opportunities were favorable, although
this is difficult to determine without age data. When
transition probabilities were calculated by gap-size
class, the only notable trend was that pole-sized
F. sylvatica showed increased gap capture in large
gaps, while the opposite trend was found for A. alba.
These results again suggest that F. sylvatica recruits
to the pole stage more successfully in large gaps,
while A. alba is better equipped to tolerate the lower
light levels at this stage in small gaps.
Patterns of canopy replacement may also provide
insight into autogenic mechanisms of coexistence
(e.g., Frelich and Reich 1995; Poulson and Platt
1996; Woods 1984). Previous workers have sug-
gested that A. alba and F. sylvatica tend to replace
each other in old-growth forests of the Dinaric
mountains (Pintaric 1978; Prpic et al. 2001). In this
study, there was no strong trend toward self- or
reciprocal replacement for F. sylvatica and A. alba at
the seedling and sapling stages. For pole-sized trees
and definitive gapfillers, we found higher probabili-
ties of reciprocal replacement, although self-replace-
ment was also relatively common. Therefore, we did
not find compelling evidence of coexistence mediated
by autogenic neighborhood effects that cause either
reciprocal or self-replacement as other studies have
demonstrated (Fox 1977; Runkle 1981; Woods 1984).
Since our observations are only a snapshot in time,
they are unlikely to capture the longer-term temporal
dynamics of forest regeneration. Indeed, relatively
short-term permanent plot studies in Carpathian
F. sylvatica–A. alba forests show that establishment
and recruitment of seedlings and saplings are highly
dynamic (Szwagrzyk and Szewczyk 2008; Szwagr-
zyk et al. 2001). Nevertheless, there is evidence that
the predominance of F. sylvatica and lack of A. alba
in the regeneration layer we observed is not a recent,
temporary pattern. Old inventories from Perucica also
documented a general pattern of abundant F. sylvat-
ica and scarce A. alba regeneration (Fukarek 1970;
Pintaric 1978). The low density of A. alba seedlings
Table 3 Transition probabilities by expanded gap-size class and the combined gap data set for seedlings, saplings, pole-sized trees,
and definitive gapfillers of the three most common gapfiller tree species
Gapmakers
a
Gapfillers
Seedlings Saplings Pole trees Definitive gapfillers
b
Fa sy Ab al Ac ps Fa sy Ab al Ac ps Fa sy Ab al Ac ps Fa sy Ab al Ac ps None
Small gaps: \250 m
2
(n=34)
Fa sy (n=19) 0.60 0.15 0.04 0.72 0.17 0 0.28 0.72 0 0.25 0.35 0 0.40
Ab al (n=27) 0.89 0.02 0.05 0.93 0 0.03 0.48 0.49 0 0.23 0.50 0 0.27
Medium gaps: 250–400 m
2
(n=24)
Fa sy (n=27) 0.76 0.04 0.17 0.83 0.02 0.15 0.22 0.78 0 0.08 0.52 0 0.40
Ab al (n=40) 0.85 0.02 0.10 0.90 0.06 0.04 0.46 0.54 0 0.21 0.33 0 0.46
Large gaps: [400 m
2
(n=24)
Fa sy (n=37) 0.82 0.07 0.10 0.97 0.01 0.02 0.63 0.35 0.01 0.33 0.41 0.06 0.20
Ab al (n=65) 0.82 0.04 0.10 0.96 0.01 0.03 0.72 0.24 0.02 0.40 0.21 0.04 0.35
All gaps (n=82)
Fa sy (n=83) 0.75 0.07 0.11 0.87 0.05 0.06 0.41 0.58 0.01 0.23 0.45 0.04 0.28
Ab al (n=132) 0.79 0.03 0.13 0.94 0.02 0.03 0.60 0.38 0.01 0.33 0.28 0.02 0.36
Fa sy:Fagus sylvatica,Ab al:Abies alba,Ac ps:Acer pseudoplatanus
a
Transition probabilities are only shown for the two most abundant gapmakers, Fagus sylvatica (Fa sy) and Abies alba (Ab al)
b
The number of gapmakers increases in the definitive gapfiller matrix for both large gaps (F. sylvatica =82; A. alba =116) and all
gaps (F. sylvatica =128, A. alba =183) because they include the five largest gaps where only definitive gapfillers were recorded
Plant Ecol (2010) 208:307–318 315
123
and saplings in Perucica, however, is difficult to
explain. Pole-sized A. alba between 10 and
20 cm dbh in Perucica are often [100 years old
(n=26, age range: 79–210; Nagel and Svoboda,
unpublished data), suggesting that the regeneration
failure of A. alba has lasted many decades. A similar
lack of A. alba regeneration reported in other F. sylv-
atica–A. alba forests in the northern Dinaric Moun-
tains is primarily attributed to high browsing pressure
by large herbivores, particularly red deer (Cervus ela-
phus L.), which reach densities [6 animals km
-2
(Klopcic et al. 2010; Rozenbergar et al. 2007).
However, red deer have been locally extirpated in
our study area due to hunting, and inventory data from
hunter surveys suggest that the two other large
herbivores present in the region, roe deer (Capreo-
lus capreolus L.) and chamois (Rupicapra rupicapra
L.), have densities far below 1 animal km
-2
(Anon-
ymous 2004). In support of this, A. alba regeneration
in Perucica does not appear to be heavily browsed.
Despite the shortcomings associated with making
observations at one point in time, we believe our
findings provide insight into tree species coexistence
that may be applicable to similar F. sylvatica–A. alba
dominated forests throughout central and southeast-
ern Europe. Here, we propose a simple conceptual
model of F. sylvatica–A. alba coexistence based on
our observations and the limited available data on the
life history characteristics of these species. Although
few studies have examined growth rates of A. alba
and F. sylvatica in different light conditions during
early life stages, there is evidence that F. sylvatica
has faster growth rates than A. alba over a range of
light conditions, particularly at higher light levels
(Burschel et al. 1985;Ha
¨ttenschwiler 2001; Preuhsler
1989, but see Stancioiu and O’Hara 2006). Further-
more, saplings and pole-sized trees of A. alba can
survive in a suppressed condition under the canopy
and slowly increase in height over very long periods
(Ferlin 2002; Mayer 1984; Szymura 2005). In
contrast, the results of our replacement probability
analysis indirectly suggest that F. sylvatica survives
for a shorter period under the canopy during the pole
stage, which could be due to a decrease in shade
tolerance with increasing plant size (Kneeshaw et al.
2006; Messier et al. 1999; Valladares and Niinemets
2008). Thus, it is possible that differences in trade-
offs between survival and growth over different light
conditions, particularly for pole-sized trees, may
promote coexistence of both species (e.g., Kobe
et al. 1995; Seiwa 2007). Variation in the frequency
and severity of gap disturbances could then promote
coexistence. For example, during periods of infre-
quent, low intensity disturbances, A. alba would
benefit because of its ability to tolerate long periods
of suppression, while periods with higher rates of
disturbance would favor F. sylvatica because of its
ability to outperform A. alba in more lit conditions.
Likewise, periodic higher severity events that create
larger multiple treefall gaps would promote less
tolerant species, such as A. pseudoplatanus,U. glab-
ra, and F. excelsior. A similar conceptual model of
allogenic coexistence of beech and maple in North
America was proposed by Poulson and Platt (1996).
Indeed, much research is needed to verify the
conceptual model presented here. A more quantita-
tive understanding of the shade tolerance of A. alba
and F. sylvatica is warranted, and in particular, how
shade tolerance shifts with ontogeny. Additionally,
retrospective studies of disturbance in stands with
varying amounts of A. alba and F. sylvatica would be
useful. It should also be noted that this model ignores
the influence of site conditions, such as soil fertility
and moisture gradients, which are likely to influence
the competitive relationships between A. alba and
F. sylvatica under different light conditions (Ammer
1996; Burschel et al. 1985). In Perucica, for example,
acidic soils that occur in parts of the reserve are
thought to provide a competitive advantage to A. alba
(Fukarek 1970). Finally, although we did not find
strong evidence of autogenic coexistence, species-
specific influences of canopy trees on regeneration
(i.e., neighborhood effects) deserve more attention.
These may include the differential influence of
A. alba and F. sylvatica canopies on understory light
transmission (Canham et al. 1994) or soil nutrients
(Finzi et al. 1998). Further investigation in these
areas would provide valuable insight into the coex-
istence of F. sylvatica and A. alba and the overall
dynamics of these forests.
Acknowledgments This study was funded by the Slovenian
Research Agency (ARRS). Miroslav Svoboda received support
from the project VaV MZP SP/2d2/111/08. We are grateful to
the park administration of Sutjeska National Park for providing
access to the Perucica forest reserve. We thank Katrine Hahn
Kristensen and Anders Busse Nielsen for help with field
sampling, and two anonymous reviewers for providing helpful
critiques of an earlier version of this manuscript.
316 Plant Ecol (2010) 208:307–318
123
References
Ammer C (1996) Konkurrenz um Licht. Zur Entwicklung der
Naturverju
¨ngung im Bergmischwald. Forstliche For-
schungsberichte Mu
¨nchen 158: 198 pp
Anonymous (2004) Game management plan for Sutjeska
National Park, Bosnia and Herzegovina
Barden LS (1981) Forest development in canopy gaps of a
diverse hardwood forest of the southern Appalachian
Mountains. Oikos 37:205–209
Battles JJ, Fahey TJ (2000) Gap dynamics following forest
decline: a case study of red spruce forests. Ecol Appl
10:760–774
Brokaw NVL (1985) Gap-phase regeneration in a tropical
forest. Ecology 66:682–687
Burschel P, Kateb H, Huss J, Mosandl R (1985) Die Verju
¨n-
gung im Bergmischwald. Forstwiss Cent bl 104: 65–100
Busing RT (1996) Estimation of tree replacement patterns in an
Appalachian Picea-Abies forest. J Veg Sci 7:685–694
Busing RT, White PS (1997) Species diversity and small-scale
disturbance in an old-growth temperate forest: a consid-
eration of gap partitioning concepts. Oikos 78:562–568
Canham CD (1989) Different responses to gaps among shade-
tolerant tree species. Ecology 70:548–550
Canham CD, Finzi AC, Pacala SW, Burbank DH (1994)
Causes and consequences of resource heterogeneity in
forests—interspecific variation in light transmission by
canopy trees. Can J For Res 24:337–349
de Ro
¨mer AH, Kneeshaw DD, Bergeron Y (2007) Small gap
dynamics in the southern boreal forest of eastern Canada:
do canopy gaps influence stand development? J Veg Sci
18:815–826
Denslow JS (1980) Gap partitioning among tropical rainforest
trees. Biotropica 12:47–55
Denslow JS (1995) Disturbance and diversity in tropical rain-
forests—the density effect. Ecol Appl 5:962–968
Dro
¨sser L, von Lu
¨pke B (2005) Canopy gaps in two virgin
beech forest reserves in Slovakia. J For Sci 51:446–457
Ellenberg HH (1988) Vegetation ecology of central Europe.
Cambridge University Press, Cambridge, UK
Ferlin F (2002) The growth potential of understorey silver fir
and Norway spruce for uneven-aged forest management in
Slovenia. Forestry 75:375–383
Finzi AC, Van Breemen N, Canham CD (1998) Canopy tree
soil interactions within temperate forests: species effects
on soil carbon and nitrogen. Ecol Appl 8:440–446
Firm D, Nagel TA, Diaci J (2009) Disturbance history and
dynamics of an old-growth mixed species mountain forest
in the Slovenian Alps. For Ecol Manag 257:1893–1901
Fox JF (1977) Alternation and coexistence of tree species. Am
Nat 111:69–89
Frelich LE, Reich PB (1995) Spatial patterns and succession in
a Minnesota southern-boreal forest. Ecol Monogr 65:325–
346
Fukarek P (1970) Sumske zajednice prasumskog rezervata
Perucica u Bosni. In: Fukarek P (ed) Simpozijum
juznoevpropske prasume i viso-koplaninska flora i vege-
tatija istocnoalpsko-dinarskog prostora. redovni clan
akademije nauka i umjetnosti Bosne i Hercegovine,
Sarajevo, pp 157–262
Fukarek P, Stefanovic V (1958) Prasuma Perucica i njena
vegetacija. Sarajevo
Givnish TJ (1988) Adaptation to sun and shade—a whole-plant
perspective. Aus J Plant Physiol 15:63–92
Grubb PJ (1977) Maintenance of species-richness in plant
communities—importance of regeneration niche. Biol
Rev 52:107–145
Gutierrez AG, Aravena JC, Carrasco-Farias NV, Christie DA,
Fuentes M, Armesto JJ (2008) Gap-phase dynamics and
coexistence of a long-lived pioneer and shade-tolerant tree
species in the canopy of an old-growth coastal temperate
rain forest of Chiloe Island, Chile. J Biogeogr 35:1674–
1687
Ha
¨ttenschwiler S (2001) Tree seedling growth in natural deep
shade: functional traits related to interspecific variation in
response to elevated CO2. Oecologia 129:31–42
Hein S, Collet C, Ammer C, Le Goff N, Skovsgaard JP, Savill
P (2009) A review of growth and stand dynamics of
Acer pseudoplatanus L. in Europe: implications for sil-
viculture. Forestry 82:361–385
Hubbell SP, Foster RB, O’Brien ST, Harms KE, Condit R,
Wechsler B, Wright SJ, de Lao SL (1999) Light-gap
disturbances, recruitment limitation, and tree diversity in a
neotropical forest. Science 283:554–557
Klopcic M, Jerina K, Boncina A (2010) Long-term changes of
structure and tree species composition in Dinaric uneven-
aged forests: are red deer an important factor? Eur J Forest
Res. doi:10.1007/s10342-009-0325-z
Kneeshaw DD, Prevost M (2007) Natural canopy gap distur-
bances and their role in maintaining mixed-species forests
of central Quebec, Canada. Can J For Res 37:1534–1544
Kneeshaw DD, Kobe RK, Coates KD, Messier C (2006) Sap-
ling size influences shade tolerance ranking among
southern boreal tree species. J Ecol 94:471–480
Kobe RK, Pacala SW, Silander JA, Canham CD (1995)
Juvenile tree survivorship as a component of shade tol-
erance. Ecol Appl 5:517–532
Lertzman KP (1992) Patterns of gap-phase replacement in a
sub-Alpine, old-growth forest. Ecology 73:657–669
Lertzman KP (1995) Forest dynamics, differential mortality
and variable recruitment probabilities. J Veg Sci 6:191–
204
Lieberman M, Lieberman D, Peralta R (1989) Forests are not
just Swiss cheese—canopy stereogeometry of non-gaps in
tropical forests. Ecology 70:550–552
Lusk CH (2004) Leaf area and growth of juvenile temperate
evergreens in low light: species of contrasting shade tol-
erance change rank during ontogeny. Funct Ecol 18:820–
828
Lusk CH, Smith B (1998) Life history differences and tree
species coexistence in an old-growth New Zealand rain
forest. Ecology 79:795–806
Mayer H (1984) Waldbau auf soziologisch-o
¨kologischer
Grundlage. Gustav Fischer Verlag, Stuttgart
Messier C, Doucet R, Ruel JC, Claveau Y, Kelly C, Lechowicz
MJ (1999) Functional ecology of advance regeneration in
relation to light in boreal forests. Can J For Res 29:812–
823
Mlinsek D (1967) Rast in sposobnost reagiranja pragozdne
bukve. Zbornik Biotehniske fakultete 15:63–79
Plant Ecol (2010) 208:307–318 317
123
Nagel TA, Diaci J (2006) Intermediate wind disturbance in an
old-growth beech-fir forest in southeastern Slovenia. Can
J For Res 36:629–638
Nagel TA, Svoboda M (2008) Gap disturbance regime in an
old-growth Fagus–Abies forest in the Dinaric Mountains,
Bosnia–Herzegovina. Can J For Res 38:2728–2737
Nagel TA, Svoboda M, Diaci J (2006) Regeneration patterns
after intermediate wind disturbance in an old-growth
Fagus–Abies forest in southeastern Slovenia. For Ecol
Manag 226:268–278
Nagel TA, Levanic T, Diaci J (2007) A dendroecological
reconstruction of disturbance in an old-growth Fagus–
Abies forest in Slovenia. Ann For Sci 64:891–897
Petritan AM, Von Luepke B, Petritan IC (2007) Effects of
shade on growth and mortality of maple (Acer pseudo-
platanus), ash (Fraxinus excelsior) and beech (Fagus
sylvatica) saplings. Forestry 80:397–412
Pintaric K (1978) Urwald Perucica als natu
¨rliches Fors-
chungslaboratorium. Allgemeine Forst Zeitschrift 24:702–
707
Piovesan G, Di Filippo A, Alessandrini A, Biondi F, Schirone
B (2005) Structure, dynamics and dendroecology of an
old-growth Fagus forest in the Apennines. J Veg Sci
16:13–28
Poulson TL, Platt WJ (1996) Replacement patterns of beech
and sugar maple in Warren Woods, Michigan. Ecology
77:1234–1253
Preuhsler T (1989) Die Entwicklung von Oberstand und Na-
turverju
¨ngung in Bergmischwald-Verju
¨ngungsbesta
¨nden
des Forstamtes Kreuth. Centralblatt fu
¨r das gesamte for-
stwesen 106:23–54
Prpic B, Matic S, Vukelic J, Seletkovic Z (2001) Virgin forests
of beech and fir in the Croatian Dinaric Mountain range.
In: Prpic B (ed) Silver Fir in Croatia. Akademija sumar-
skih znanosti—Hrvatske sume, Zagreb, pp 493–500
Rebertus AJ, Veblen TT (1993) Structure and tree-fall gap
dynamics of old-growth Nothofagus forests in Tierra-Del-
Fuego, Argentina. J Veg Sci 4:641–654
Rozenbergar D, Mikac S, Anic I, Diaci J (2007) Gap regen-
eration patterns in relationship to light heterogeneity in
two old-growth beech-fir forest reserves in South East
Europe. Forestry 80:431–443
Runkle JR (1981) Gap regeneration in some old-growth forests
of the eastern United States. Ecology 62:1041–1051
Runkle JR (1982) Patterns of disturbance in some old-growth
Mesic Forests of Eastern North America. Ecology 63:
1533–1546
Schnitzer SA, Carson WP (2001) Treefall gaps and the main-
tenance of species diversity in a tropical forest. Ecology
82:913–919
Seiwa K (2007) Trade-offs between seedling growth and sur-
vival in deciduous broadleaved trees in a temperate forest.
Ann Bot 99:537–544
Stancioiu PT, O’Hara KL (2006) Regeneration growth in dif-
ferent light environments of mixed species, multiaged,
mountainous forests of Romania. Eur J For Res 125:151–
162
Stewart GH, Rose AB, Veblen TT (1991) Forest development
in canopy gaps in old-growth beech (Nothofagus) forests,
New Zealand. J Veg Sci 2:679–690
Szewczyk J, Szwagrzyk J (1996) Tree regeneration on rotten
wood and on soil in old-growth stand. Vegetatio 122:37–
46
Szwagrzyk J, Szewczyk J (2008) Is natural regeneration of
forest stands a continuous process? A case study of an old-
growth forest of the western Carpathians. Pol J Ecol
56:623–634
Szwagrzyk J, Szewczyk J, Bodziarczyk J (2001) Dynamics of
seedling banks in beech forest: results of a 10-year study
on germination, growth and survival. For Ecol Manag
141:237–250
Szymura TH (2005) Silver fir sapling bank in seminatural
stand: individuals architecture and vitality. For Ecol
Manag 212:101–108
Taylor AH (1990) Disturbance and persistence of Sitka spruce
(Picea-sitchensis (Bong.) Carr.) in coastal forests of the
pacific northwest, North America. J Biogeogr 17:47–58
Uhl C, Clark K, Dezzeo N, Maquirino P (1988) Vegetation
dynamics in Amazonian treefall gaps. Ecology 69:
751–763
Valladares F, Niinemets U (2008) Shade tolerance, a key plant
feature of complex nature and consequences. Annu Rev
Ecol Evol Syst 39:237–257
Veblen TT (1986) Treefalls and the coexistence of conifers
in sub-alpine forests of the central Rockies. Ecology
67:644–649
Webb SL, Scanga SE (2001) Windstorm disturbance without
patch dynamics: twelve years of change in a Minnesota
forest. Ecology 82:893–897
White PS, Mackenzie MD, Busing RT (1985) A critique on
overstory understory comparisons based on transition-
probability analysis of an old growth spruce-fir stand in
the Appalachians. Vegetatio 64:37–45
Whitmore TC (1989) Canopy gaps and the two major groups of
forest trees. Ecology 70:536–538
Woods KD (1984) Patterns of tree replacement: canopy effects
on understory pattern in hemlock-northern hardwood
forests. Vegetatio 56:87–107
318 Plant Ecol (2010) 208:307–318
123
