Neighbourhood-based evidence of tree diversity promotion by beech in an old-growth deciduous-coniferous mixed forest (Eastern Carpathians)

Page created by Jorge Malone
 
CONTINUE READING
Neighbourhood-based evidence of tree diversity promotion by beech in an old-growth deciduous-coniferous mixed forest (Eastern Carpathians)
Ann. For. Res. 64(1): 13-30, 2021                                         ANNALS OF FOREST RESEARCH
https://doi.org/10.15287/afr.2021.2143				                                www.afrjournal.org

Neighbourhood-based evidence of tree diversity
promotion by beech in an old-growth deciduous-
coniferous mixed forest (Eastern Carpathians)

Dan Gafta1 , Annik Schnitzler2, Déborah Closset-Kopp3, Vasile Cristea1
           @

                       Gafta D., Schnitzler A., Closset-Kopp D., Cristea V., 2021. Neighbourhood-based
                       evidence of tree diversity promotion by beech in an old-growth deciduous-coniferous
                       mixed forest (Eastern Carpathians). Ann. For. Res. 64(1): 13-30.
                       Abstract Neighbourhood models are useful tools for understanding the role
                       of positive and negative interactions in maintaining the tree species diversity
                       in mixed forests. Under such a presumption, we aimed at testing several
                       hypotheses concerning the mechanisms of autogenic species coexistence in
                       an old-growth, beech-fir-spruce stand, which is part of the Slătioara forest
                       reserve (Eastern Carpathians). Univariate/bivariate spatial point pattern
                       analyses, the individual tree species-area relationship, the species mingling
                       analysis and generalised linear mixed models of neighbour interference
                       were applied on data concerning the position and allometry of all saplings
                       and trees occurring within a 0.24 ha plot.
                       The monospecific distribution of either beech or spruce saplings did
                       not support the spatial segregation hypothesis. There was no evidence
                       of conspecific negative distance dependence, as no spatial segregation
                       was detected between the saplings and trees of any species. Within 4
                       m-neighbourhood, the beech saplings appeared as diversity accumulators,
                       which might be indicative of indirect facilitation (e.g., herd protection
                       hypothesis). At tree stage, none of the three species showed either
                       accumulator or repeller patterns in their neighbourhood with respect to
                       sapling species richness. Signals of positive and negative interspecific
                       association were found in tree-sized beech (at scales of 10 to 20 m) and
                       spruce (at scales of 4 to 17 m), respectively. The former, highly interspersed
                       pattern is in accordance with the hypothesis of positive complementary
                       effects, whereas the latter, poorly intermingled pattern is probably linked
                       to the unexpected, positive neighbouring effect of spruce trees on the stem
                       growth of their conspecific saplings. Such self-favouring process might
                       be due to a facilitative below-ground mechanism. Conversely, the beech
                       saplings were suppressed through interference from the neighbouring
                       conspecific trees.
                       The beech appears to be the key promoter of tree species coexistence in
                       the study forest stand, in contrast to the low interspersion of spruce in the
                       overstorey leading to lower local tree diversity.
                       Keywords: auto-facilitation, beech-fir-spruce mixed forest, diversity
                       accumulator, neighbourhood effect, null models, self-competitor, spatial
                       point pattern, tree species mingling.

                                                                                                       13
Neighbourhood-based evidence of tree diversity promotion by beech in an old-growth deciduous-coniferous mixed forest (Eastern Carpathians)
Ann. For. Res. 64(1): 13-30, 2021                                                                 Research article

                             Addresses: 1‌Department of Taxonomy and Ecology, Centre 3B, Babeș-Bolyai
                             University, Romania‌| 2Laboratoire Interdisciplinaire des Environnements
                             Continentaux, Université de Lorraine, Metz, France| 3 Unité Ecologie et Dynamique
                             des Systèmes Anthropisés, Université de Picardie Jules Verne, Amiens, France.
                              @
                                    Corresponding Author: Dan Gafta (dan.gafta@ubbcluj.ro).

                             Manuscript received February 8, 2021; revised May 14, 2021; accepted May 26, 2021.

Introduction                                                 their close neighbours. Such interactions
                                                             may lead to either negative or positive net
Along with propagule availability and                        effects, mainly depending on the balance
environmental filters, niche-based interactions              between asymmetric competition for limiting
have been widely considered as the main                      resources (light, nutrients) and facilitation
drivers of species coexistence and community                 (nursering), i.e. either resource limitation or
assembly (Wilson 2011, Kraft et al. 2015). In                diversification (Chi et al. 2015, Bulleri et al.
forest communities, a number of mechanisms                   2016). Other possible mechanisms underlying
have been proposed to explain the local scale-               the negative and positive interactions between
coexistence of tree species: different patterns              neighbouring plants may include shared pests
in seed dispersal and persistence (Harms et al.              (e.g., pathogens, herbivores) and respectively,
2001, Wright 2002, Hou et al. 2004, Punchi-                  shared mutualists - like mycorrhizae and seed
Manage et al. 2015); habitat heterogeneity                   dispersers (Punchi-Manage et al. 2015).
(Koukoulas & Blackburn 2005, King et al.                       In temperate mixed forests composed of more
2006, Getzin et al. 2008, Chen et al. 2010,                  or less shade-tolerant tree species, dominance
Piao et al. 2013); autogenic alternation                     can easily change from one species to another
(reciprocal replacement) of species between                  between consecutive generations (Woods
ontogenetic cycles (Fox 1977, Woods 1979);                   1984, Arii & Lechowicz 2002). Negative
allogenic formation of canopy gaps by natural                spatial association between conspecifics by
disturbance (Canham 1988, Poulson & Platt                    means of reciprocal replacement were reported
1996) and demographic stochasticity (Gravel                  in mixed deciduous-coniferous forests from
et al. 2008). Recent studies have suggested                  different geographic regions (Bândiu 1977,
that niche structuring is a major determinant                Nakashizuka & Kohyama 1995, Akashi
of species diversity in temperate mixed forests              1996, Kuninaga et al. 2015). The autogenic
(Gilbert & Lechowicz 2004, Laliberté et al.                  coexistence through reciprocal replacement
2009, Zhang et al. 2010). Two different classes              of tree species can originate from various
of biotic mechanisms mediate coexistence by                  processes like, direct facilitation (e.g., light
preventing competitive exclusion (Chesson                    spectral filtering), intraspecific inhibition (e.g.,
2000): equalising mechanisms, which reduce                   auto-allelopathy) or indirect facilitation (e.g.,
the relative fitness difference between species              ‘escape’ and ‘herd immunity’ hypotheses).
(e.g., facilitation or symmetric competition)                As a necessary but not sufficient condition, it
and stabilising mechanisms, which reduce                     seems that coexistence requires interspecific
niche overlap (e.g., specialisation or self-                 differences in light transmissivity through the
inhibition).                                                 crowns of adult trees (Cammarano 2011). For
  The niche-based mechanisms related to                      instance, fir regenerates better under beech
spatial processes are especially important for               canopy due to the greater transmission of blue
plant species coexistence because they are                   and red light (Bândiu 1977, Dobrowolska
sessile organisms that interact mainly with                  1998). On the contrary, the self-inhibition of
14
Gafta et al.                                                Neighbourhood-based evidence of tree diversity ...

fir saplings through auto-intoxication has been          interactions can cause local maxima and
documented in pure silver fir stands (Becker &           respectively, minima in species richness, the
Drapier 1984, 1985).                                     analysis of individual species-area relationship
   The ‘escape hypothesis’ or ‘conspecific               (ISAR) may reveal the taxonomic identity of
negative density/distance-dependence’ (CNDD)             the so-called diversity ‘accumulators’ and
assumes that, because the dispersed seed density         respectively, ‘repellers’ (Wiegand et al. 2009).
is typically highest near the parent tree, specialised   Finally, the third category includes the so-
enemies (pathogens, herbivores) accumulate and           called ‘neutral species’, which do not display
reduce seedling establishment near conspecific           patterns (peaks or saddles) in neighbourhood
trees, resulting in lower intraspecific aggregation      species richness and are presumably related
(Comita et al. 2014). Recent studies brought             to stochastic assorting (Wiegand et al.
evidence of CNDD as being an important                   2007). Species displaying high-diversity
driver of species coexistence in temperate old-          neighbourhoods may indicate a preponderance
growth forests (Getzin et al. 2008, Johnson et           of positive interspecific interactions or strong
al. 2014, Kuang et al. 2017), including mixed            CNDD, while species with low-diversity
coniferous-deciduous forests (Nakashizuka                neighbourhoods may indicate dominance of
& Kohyama 1995, Hiura & Fujiwara 1999,                   inhibitory effects on heterospecific neighbours
Kotanen 2007, Bai et al. 2012, Piao et al. 2013).        or strong conspecific positive density-
The ‘herd immunity hypothesis’ gives a similar           dependence. As a consequence, the proportion
explanation of the reciprocal replacement                of diversity ‘repellers’ and ‘accumulators’
of trees species but from a different angle: it          within a community could shed light on
predicts that high local species diversity confers       the mechanisms ruling coexistence in tree
protection from natural enemies by rendering             communities (Espinosa et al. 2015).
it more difficult for specialist natural enemies            The ‘spatial segregation hypothesis’, which
to locate the target plants (Wills et al. 1997).         involves negative spatial association between
If both the escape and herd immunity theories            heterospecific individuals, can also be related
hold, there should be a tendency toward high             to species coexistence (Pacala & Levin 1997).
spatial mingling of tree species and ultimately,         Intraspecific aggregation of juveniles (due to
a high local species diversity (Blundell & Peart         limited seed dispersal or gap-based regeneration
2004, Volkov et al. 2005, Swamy et al. 2011,             niche) leads to interspecific segregation, which
Pommerening & Uria-Diez 2017).                           in turn prevents the exclusion of competitively
   The approach based on distance­-dependent             inferior species (Stoll & Prati 2001), thereby
analyses can be also employed to assess                  promoting species diversity. At later life
the performance of a focal tree based on the             stages, stable coexistence of tree species can
characteristics of its nearest neighbours.               be achieved by ‘positive complementary
Usually, the target tree growth and/or survival          effects’ that arise owing to great niche
is analysed as a function of the size and                differentiation (Cavard et al. 2011, Lasky et
distance from the neighbours (Wagner &                   al. 2014, Forrester & Bauhus 2016). Several
Radosevich 1998, Canham et al. 2004, Kunstler            mechanisms may be responsible for such
et al. 2016), but the taxonomic identity of the          effects, like reduction in crown interference
latter has been shown to be equally important            due to spatial stratification (Pretzsch 2014)
(Uriarte et al. 2004). If the net effect of the          or improved nutrient availability by virtue of
nearest neighbours on the focal trees is not null,       more efficient exploitation of soil resources
then spatial structure in local species richness         (Rothe & Binkley 2001).
may emerge (Lieberman & Lieberman 2007).                    The main goal of this study was to search
Since positive and negative interspecific                for spatial dependence and neighbourhood

                                                                                                           15
Ann. For. Res. 64(1): 13-30, 2021                                                     Research article

patterns across/within tree species and size        (Eastern Carpathians, Romania) between 800
classes in an old-growth, mixed beech-fir-          and 1510 m above sea level (Fig. 1). The
spruce forest, in order to test hypotheses about    climate is temperate-continental with mean
the underlying mechanisms of autogenic              annual temperatures between 3.9 and 5.8°C
species coexistence. We hypothesized that           and mean annual precipitation ranging from
local tree species diversity could be promoted      700 to 810 mm (Duduman et al. 2014). The
and maintained by: i) facilitative interactions     forest reserve was created in 1941 and covered
between heterospecific saplings in the shaded       an area of 854.3 ha until in 2006, when it
understorey; ii) effects of conspecific negative    was extended to 1064.2 ha. The mixed forest
distance-dependence, self-inhibition and            canopy is formed of European beech (Fagus
positive interspecific interactions between trees   sylvatica), silver-fir (Abies alba) and Norway
and saplings; iii) positive complementarity         spruce (Picea abies), except in the upper
effects between canopy trees.                       montane stands (over 1350 m of elevation) that
   Assuming       that     such     deterministic   are beyond the beech altitudinal range.
mechanisms of autogenic species coexistence            The investigations were undertaken on a
exert stronger effects than negative                mild (25%) north-western slope covered by
interspecific interactions in self-regenerating,    rendzina soils, at an elevation of about 830
stable, old-growth mixed forests, we expected       m a.s.l. By reference to living plants only,
to observe (at small spatial scales) patterns of    the beech was by far the most abundant tree
aggregation/segregation or positive/negative        species in both over- and under-storey (67%),
association between heterospecific/conspecific      followed by spruce (24%) that rarely reached
individuals, and negative neighbourhood             the uppermost level of the forest canopy (Table
effects of trees on the growth of conspecific       1). The fir had a modest share of 9% and was
saplings. By reference to appropriate null          poorly represented in the understorey, in spite
models of spatial point pattern, we analysed:       of the fir trees having cumulated the largest
i) the distribution of conspecific saplings for     basal area and being on average the tallest
testing the spatial segregation hypothesis; ii)     among the three species (Table 1). In terms
the distribution of saplings of each species        of proportion of dead standing individuals in
with respect to their conspecific/heterospecific    each species population, spruce was ranked
trees for testing the reciprocal replacement        first (22.7%), followed by fir (9.8%) and beech
hypothesis, and iii) the individual tree species-   (only 4.8%). Overall, the tree population
area relationship and tree species mingling         structure was uneven but also heterogeneous
for testing the diversity accumulator/repeller      throughout the Slătioara forest (Cenușă et
hypothesis and respectively, the positive           al. 2002), and in particular, the study stand
complementarity effects. In addition, we            featured a low number of trees in the middle
estimated the effect of the nearest conspecific/    size classes (Schnitzler et al. 2004). According
heterospecific tree size on the height of           to the forest classification on ecosystemic
focal saplings for testing the hypothesis of        bases by Doniță et al. (1990), the study stand
asymmetric competition.                             can be assigned to the forest ecosystem type
                                                    2316, characterised by highly to moderately
Materials and Methods                               productive spruce-beech-fir stands, developed
                                                    on eu-mesobasic brown or rendzinic soils, well
Study area                                          balanced in terms of water supply, featuring
                                                    a mull-moder humus type and a herbaceous
The Codrul Secular Slătioara forest reserve         layer of Oxalis-Dentaria-Asperula type.
(47°27′N; 25°37′E) lies in the Rarău Mountains

16
Gafta et al.                                               Neighbourhood-based evidence of tree diversity ...

Figure 1 Geographic location of the Codrul Secular Slătioara forest reserve (left) and the approximative
         placement of the inventoried plot within the protected area (right).

Table 1 Summary statistics of the population size and allometry of the living plants (seedlings excluded) by
        tree species in the study forest stand.
Variable                  Statistics                       Fir               Beech              Spruce
Density (N/ha)            Sum                             154                 1154               413
Basal area (m2/ha)        Sum                             31.9                17.8                8.2
                          Range (min - max)            3.2 - 114.5         3.2 - 79.3         3.2 - 116.8
Diameter (cm)             Median                          28.1                 4.8                6.3
                          Mean ± SD                    42.9 ± 38.2         13.5 ± 16.9       10.3 ± 17.4
                          Range (min - max)            1.0 - 48.0          0.1 - 38.0         0.9 - 51.5
Height (m)                Median                          12.5                 3.7                2.9
                          Mean ± SD                    18.0 ± 16.9          6.4 ± 8.3          4.4 ± 7.2

Data collection                                        clinometer and respectively, a tape ruler. In
                                                       addition, the stem girth at breast height (1.30
The field work was carried out in 2001 within          m) was measured on all individuals that
a 40 x 60 m rectangular plot circumscribed             were at least 1.50 m tall. Two size classes
to a relatively homogeneous area in terms of           were distinguished based on trunk thickness,
site conditions and stand physiognomy, and             i.e. individuals displaying a stem girth of at
located (at that time) within the buffer zone of       least 10 cm (DBH ≥ 3.2 cm) were considered
the forest reserve. The long sides of the plot         trees, whereas all others were included in the
were placed along the steepest slope path, i.e.        category of saplings.
perpendicular to the contour lines.
   The height and Cartesian coordinates                Data analysis
(with respect to one plot corner) of all living
and dead standing tree species individuals             We used point pattern analyses of fully
(except seedlings) were measured using a               mapped plant locations to explore the

                                                                                                            17
Ann. For. Res. 64(1): 13-30, 2021                                                         Research article

univariate and bivariate spatial patterns in          distributions or statistics (i.e., g(r), ISAR
the distribution of individuals (Wiegand &            and M) was assessed through simulations of
Moloney 2014). The univariate analysis was            (density constrained) random distribution of
employed on conspecific saplings to test the          trees or saplings, in order to account for possible
spatial segregation hypothesis. The bivariate         (undesirable) effects of habitat patchiness,
analysis was performed on saplings relative           stand history and dispersal limitation that can
to their conspecific or heterospecific trees, in      induce non-random patterns in the spatial
order to test the auto-inhibition or the CNDD         disposal of individuals, irrespective of biotic
hypothesis, respectively. In the spatial point        interactions (Wiegand et al. 2007, Rayburn &
pattern analysis we used the pair-correlation         Wiegand 2012, Baddeley et al. 2014, Espinosa
function g(r), which is a second-order statistic      et al. 2015, Tsai et al. 2015). To account
that employs the probability of observing             for large variation in tree density within the
a pair of points separated by a distance r            plot, we only used inhomogeneous Poisson
(Illian et al. 2008). We computed the g(r)            processes along with the Ripley’s isotropic
distributions applying a lag distance of 0.5 m        edge correction in simulating the null models
over a maximum range of 20 m (i.e., half the          of spatial independence. The tree or sapling
minimum side of the plot).                            density throughout the plot was computed non-
   Neighbourhood effects on tree species              parametrically by employing the Epanechnikov
richness and mingling were tested using               kernel estimators of the intensity function with
species-by-all-species analyses, which integrate      the option for Jones-Diggle improved edge
across all species around a focal species             correction (Diggle 2010, Wiegand & Moloney
and are more likely to detect effects of biotic       2014).
interactions than species-by-species analyses            In the univariate g(r) analysis, the null
(Punchi-Manage et al. 2015). The individual           hypothesis was simulated through the random
species-area relationship (ISAR) estimates the        (but density-constrained) spatial disposal
change in species richness with increasing area       of conspecific saplings (Baddeley et al.
and distance from heterospecific neighbours           2015). The null model used for testing the
to target species individuals by integrating the      significance of the bivariate g(r) and ISAR
spatial structure of individuals (Wiegand et al.      was the independence of the two spatial
2007, Wiegand & Moloney 2014). Depending              point pattern types (across size classes and/
on the occurrence or absence of local minima or       or species), i.e. no spatial interaction between
maxima of species richness with respect to a null     them (Baddeley et al. 2015). The simulations
model, the ISAR approach allows the grouping          corresponding to the last null model were
of target species in (diversity) accumulators,        performed by randomly reallocating the
repellers and neutrals (Wiegand et al. 2007). The     spatial position of neighbouring saplings
mingling index (M) is a measure of interspersion      with punctual probabilities derived from the
of trees of different species and is defined as the   estimated density (intensity), while keeping
proportion of the n nearest neighbours that do        the spatial arrangement of saplings or trees
not belong to the same species as the reference       of focal species fixed (Wiegand & Moloney
tree (Pommerening & Grabarnik 2019). If the           2014). Finally, the significance of the observed
nearest neighbours and the target tree always         M index values was tested through the null
share the same species, then M = 0 (minimum           model of random labelling (independence
intermingling), else if all neighbours are always     between marks and points), that is by shuffling
taxonomically different from the target tree          the taxonomic identity of trees while keeping
species, then M = 1 (maximum intermingling).          the number and spatial position of trees in
   The significance of all previously mentioned       each species constant (Baddeley et al. 2015).

18
Gafta et al.                                          Neighbourhood-based evidence of tree diversity ...

Two-sided, pointwise envelopes for the             analyses, as the allometric measures were not
observed g(r), ISAR and M distributions were       involved in computations. On the contrary,
generated from 999 Monte Carlo simulations         only living individuals were considered in the
and an overall goodness-of-fit (GoF) test was      neighbourhood interference analysis through
performed (Loosmore & Ford 2006).                  GLMMs.
  To describe the neighbourhood interference         Because the total number of fir saplings was
of trees on the target sapling, we employed the    too low (seven living and one dead), the results
Weiner’s (1984) measure (W):                       of analyses involving their spatial point pattern
                      n                            exclusively or their allometric characteristics
                            si
               W                                 were either incomplete or biased, and hence
                     i 1   di2                    were not reported.
where n is the total number of neighbours, di        All numerical analyses were performed in R
is the distance from the target sapling to the     v3.6.3 environment (R Core Team 2020) using
i-th neighbouring tree and si is the size of the   the packages ‘spatstat’ (Baddeley et al. 2020),
i-th neighbouring tree. The size of trees was      ‘idar’ (de la Cruz 2019) and ‘spatialsegregation’
roughly estimated by reference to the volume       (Rajala 2019), except for GLMMs that were
of a cone whose dimensions were equal to the       run in SAS/STAT 9.4 (SAS Institute Inc. 2014).
tree’s diameter and height. We calculated the
W values for each neighbouring tree species        Results
separately by considering only the nearest
conspecific and heterospecific trees (n=1). In     A weak but significant aggregation of beech
the rare cases in which the focal sapling was      saplings was detected only at 1 m-scale (Fig.
taller than one of the neighbouring tree, the      2a). At larger scales, the distribution of beech
corresponding W measure was equalled to            saplings was not significantly different from the
zero.                                              null model. The spruce saplings were randomly
   Generalised     linear    mixed       models    distributed at all scales (Fig. 2b).
(GLMMs) were employed to analyse the                  Saplings of any species did not show
sapling height in each species as a function       significant spatial patterns with respect to their
of the interference from each of the nearest       conspecific (Fig. 3a-c) or heterospecific trees
neighbouring, conspecific and heterospecific       (Fig. 3d-i).
trees, while controlling the influence of             Within 4 m-neighbourhood of beech saplings,
spatial autocorrelation (Bolker et al. 2009).      a higher taxonomic sapling richness than
The latter was handled through a residual          expected under the null hypothesis was detected
random component with an anisotropic power         (Fig. 4a). Instead, the individual species-area
covariance structure, which provided the best      relationships (ISARs) corresponding to the
fitting results and could account for possible     fir and spruce saplings were not significantly
differences induced by the terrain slope. The      different from their simulated counterparts
negative binomial distribution along with a log-   (Fig. 4b-c). When the trees and saplings were
link was employed in GLMMs for adjusting           considered as focal and respectively, target
the conditional probability distribution of the    individuals, the ISARs built for the three species
response variable (i.e., sapling height). The      (beech, fir and spruce) were all fully embedded
goodness-of-fit of each model was assessed         within the corresponding simulation envelopes
through the generalised chi-square divided by      (Fig. 4d-f).
the degrees of freedom (Gen. Chi-sq/DF).              A significant, positive, spatial association
   Both living and dead individuals were           between trees species was detected at scales
considered in g(r), ISAR and species mingling      larger than 10 m around the beech trees, as indicated

                                                                                                     19
Ann. For. Res. 64(1): 13-30, 2021                                                                                    Research article

by the large observed values of the mingling                     Table 2 Standardised coefficients associated with
index compared to the simulated values (Fig.                             the (fixed) interference effects (W) of the
5a). An opposite pattern was detected within                             nearest neighbour beech, fir and spruce
a radius of 4 to 17 m around the spruce trees,                           tree on the height of beech (n=156) and
that is a significant, negative, spatial association                     spruce (n=37) saplings. The goodness-
between the three species (Fig. 5b). Finally, no                         of-fit of each GLMM is reported in the
significant spatial association between the three                        last row
species was observed in the neighbourhood of
fir trees (Fig. 5c).                                                       Response                      Sapling height
   The height of beech saplings was significantly,
negatively affected by the interference from the                           Predictors           Beech                Spruce
nearest conspecific tree, whereas the negative
                                                                 W (beech trees)                -1.282 **            -0.372 ns
effects of the nearest fir and spruce tree were not
significant (Table 2). On the contrary, the nearest              W (fir trees)                  -0.522 ns            -0.006 ns
spruce tree displayed a significant, positive                    W (spruce trees)               0.084 ns             0.494 *
effect on the height of conspecific saplings
                                                                 Intercept                      0.992                0.446
(Table 2). The nearest beech exerted a non-
significant, negative effect on the spruce sapling               Gen. Chi-sq/DF                 1.00                 1.10
height, whereas the neighbourhood effect of the                  ** 0.001
Gafta et al.                                                                                        Neighbourhood-based evidence of tree diversity ...

                                                                     10
                                                                                                      b Fir saplings

                                                                                                                               4
                                                                                                                                             c Spruce saplings vs Spruce trees
         3.0

                   a Beech saplings vs Beech trees                                                           vs
                                                                                                         Fir trees

                                                                     8
         2.5

                                                                                                                               3
g12(r)

                                                                     6
         2.0

                                                                                                                               2
                                                                     4
         1.5

                                                                                                                               1
                                                                     2
         1.0

                                                                     0
         0.5

                                                                                                                               0
               0     5             10             15          20          0         5        10               15          20       0             5        10          15          20
                                           Scale radius (m)                                       Scale radius (m)                                             Scale radius (m)

                                                                   d Fir saplings                                      e Fir saplings vs Spruce trees
                                                                           vs
                                  5

                                                                     Beech trees

                                                                                                     8
                                  4
                          g12(r)

                                                                                                     6
                                  3

                                                                                                     4
                                  2

                                                                                                     2
                                  1
                                  0

                                                                                                     0

                                       0               5       10              15       20                0           5                10            15         20

                                                                     f Beech saplings
                                                                                                                   g Beech saplings vs Spruce trees
                                                                                                    1.5

                                                                             vs
                             8

                                                                         Fir trees
                             6
                         g12(r)

                                                                                                    1.0
                             4

                                                                                                    0.5
                             2
                             0

                                                                                                    0

                                   0               5          10              15        20                0           5                10            15         20

                                                                   h Spruce saplings
                             8

                                                                                                                    i Spruce saplings vs Beech trees
                                                                                                    3.0

                                                                           vs
                                                                        Fir trees
                                                                                                    2.5
                             6
                         g12(r)

                                                                                                    2.0
                             4

                                                                                                    1.5
                                                                                                    1.0
                             2

                                                                                                    0.5
                             0

                                                                                                    0

                                   0               5          10              15        20                0           5                10            15         20
                                                                    Scale radius (m)                                                        Scale radius (m)

Figure 3 Empirical distribution of the bivariate pair-correlation function g12(r) associated with the
        spatial distribution of beech, fir and spruce saplings with respect to their conspecific (a-c) and
        respectively, heterospecific (d-i) trees. Simulation envelopes (in grey) as in Fig. 1.

                                                                                                                                                                                  21
Ann. For. Res. 64(1): 13-30, 2021                                                                                                                        Research article
                    3

                                u = 0.0602

                                                                            3

                                                                                                                                  3
                                p = 0.001
  Number of species

                                                a Focal beech saplings
                    2

                                                                            2

                                                                                                                                  2
                                                                                                         b Focal fir saplings                       c Focal spruce saplings

                                                                            1

                                                                                                                                  1
                    1
                    0

                                                                            0

                                                                                                                                  0
                        0             5           10        15         20         0       5          10           15         20         0   5          10           15          20
                    4

                                                                                                                                  4
                                                                            3
Number of species

                                                   d Focal beech trees

                                                                                                                                  3
                    3

                                                                                                                                                       f Focal spruce trees
                                                                                                            e Focal fir trees
                                                                            2

                                                                                                                                  2
                    2

                                                                            1

                                                                                                                                  1
                    1
                    0

                                                                                                                                  0
                                                                            0

                        0             5           10        15         20         0   5            10           15          20          0   5          10      15          20
                                           Neighbourhood radius (m)                       Neighbourhood radius (m)                          Neighbourhood radius (m)

   Figure 4 Observed individual species-area relationships (ISARs) built within-saplings (a-c) and across-
            size classes (d-f) by using every possible focal species. The number of species on the Y axis
            includes the focal species (size class). The statistics (u) of the goodness-of-fit test is displayed for
            significance validation. Simulation envelopes (in grey) as in Fig. 1.
                                                                                                                                  1.0

                                                                                                                                                     c Focal spruce trees
                                                                            1.0

                                                                                                        b Focal fir trees
                      0.6

                                                                                                                                  0.9
Mingling index

                                                                                                                                  0.8
                                                                            0.9
                      0.5

                                                                                                                                  0.7
                                                                            0.8

                                                          u = 0.1457
                                                                                                                                  0.6

                                                          p = 0.002
                      0.4

                                                                                                                                  0.5

                                                                                                                                                                    u = 0.2559
                                                                            0.7

                                                                                                                                                                    p = 0.002
                                                                                                                                  0.4

                                                   a Focal beech trees
                      0.3

                                                                                                                                  0.3
                                                                            0.6

                            0          5          10        15         20         0       5          10          15          20         0   5          10           15          20
                                           Neighbourhood radius (m)                           Neighbourhood radius (m)                          Neighbourhood radius (m)

   Figure 5 Empirical distribution of the mingling index (M) by neighbourhood radius built for trees only by
            using the beech (a), fir (b) and spruce (c) as focal species. The statistics (u) of the goodness-of-fit
            test is displayed for significance validation. The grey area represents a 99.9% simulation envelope
            based on the null model of random labelling. The envelopes are truncated at large scales because
            of the null variance (equal observed and simulated values of the mingling index).

    22
Gafta et al.                                         Neighbourhood-based evidence of tree diversity ...

Discussion                                            In spite of the well-documented process
                                                   of reciprocal replacement between beech
The monospecific distribution of either beech      and fir in mixed forests (Bândiu 1977, Heiri
or spruce saplings did not support the spatial     et al. 2009, Vrška et al. 2009, Diaci et al.
segregation hypothesis, which was found to         2010, Nagel et al. 2010), we did not observe
explain the tree species coexistence in other      advance regeneration of beech/fir saplings in
temperate mixed forests (e.g., Wang et al.         the neighbourhood of fir/beech trees, which is
2010, Zhou et al. 2019). Our results are, to       in accordance with the findings from southern
some extent, unexpected as both beech and          Carpathians (Petrițan et al. 2015). Similar
spruce are known to regenerate regularly           independent spatial distributions between trees
within canopy gaps (Nagel et al. 2006, Paluch      and saplings were also revealed at the other two
et al. 2019). Beech is a typical gap-filler        pairs of species: spruce/beech and fir/spruce.
capable of rapid crown enlargement (Nagel et       Possibly, the demographic heterogeneity and
al. 2010, Janík et al. 2016), whereas spruce is    stochasticity may have distorted the ecological
an early and competitive coloniser of canopy       response at species level (Hurtt & Pacala
openings (Jonášová & Prach 2004).                  1995) and may have diluted the spatial patterns
   We did not find evidence of conspecific         in tree species distribution making them weak
negative distance/density dependence (CNDD)        or undetectable (Gravel et al. 2008). On the
as no spatial segregation was detected between     other hand, a number of studies documented
saplings and trees of any species. However,        the positive effects of beech trees on fir
Janík et al. (2014) and Petrițan et al. (2015)     regeneration (Dobrowolska & Veblen 2008,
reported spatial segregation between juvenile      Vrška et al. 2009, Paluch et al. 2016, Paluch et
and mature beech individuals in old-growth         al. 2019).
beech-fir forests in western and southern             At sapling stage, only beech displayed
Carpathians, respectively. Likewise, Kuninaga      significant      diversity    patterns    in    its
et al. (2015) and Ramage & Mangana (2017)          neighbourhood. The apparent role of diversity
found evidence of CNDD in Fagus crenata            accumulator of beech saplings cannot be
growing in mixed conifer-hardwood forests and      attributable to segregation as such a pattern
respectively, in Fagus grandifolia-dominated,      was not revealed in the distribution of beech
mixed hardwoods. Unfortunately, we could           recruits. Higher than expected tree species
not test for CNDD in fir distribution because      richness in the neighbourhood of beech saplings
of the low number of saplings. Many studies        is in accordance with the positive, although
reported the decline of fir populations in mixed   weak, correlations between seedling densities
stands due to poor seedling recruitment and        in the beech-fir and beech-spruce mixtures, as
sapling survival (Diaci et al. 2011, Szwagrzyk     reported by Paluch et al. (2019), and might
et al. 2012, Paluch & Jastrzębski 2013, Janík      be indicative of net positive interactions or
et al. 2014, Keren et al. 2014, Parobeková         favourable microsites (e.g., gaps) for multiple
et al. 2018). It seems that not plant-plant        species regeneration (Wiegand et al. 2007,
interactions are responsible for the regression    Espinosa et al. 2015). Due to their relatively
of fir in mixed stands, but rather a series of     small size, it is unlikely that the beech saplings
other factors among which high accumulation        have acted as nurse plants (direct facilitation)
of beech litter (Simon et al. 2011, Janík et al.   with respect to fir and spruce saplings, but
2014) and damage by ungulates (Vrška et al.        some sort of indirect facilitation cannot be
2009, Klopcic et al. 2010). However, none of       excluded. For instance, a plausible explanation
these factors seemed to have been acting in the    could be the herd protection (immunity) theory,
study forest area.                                 i.e. enhanced survival linked to a reduced risk

                                                                                                    23
Ann. For. Res. 64(1): 13-30, 2021                                                      Research article

of transmission of species-specific pests and       Bolte et al. (2013) demonstrated that beech
pathogens because of the reduced intraspecific      can adopt a flexible root foraging strategy
density in more diverse assemblages (Peters         to access soil resources less exploited by
2003, Comita et al. 2010).                          spruce, which instead maintains a conservative
   At tree stage, none of the three species         strategy by keeping a shallow vertical fine root
showed either accumulator or repeller patterns      distribution. Such positive complementary
in their neighbourhood with respect to sapling      effects are partly responsible for the tendency
species richness. The fact that beech trees         of large trees towards high species mingling,
do not act as diversity accumulators, like          as revealed in several mixed forests across
their juveniles, may be caused by the shift in      Europe (Pommerening & Uria-Diez 2017).
habitat preferences during their ontogenetic        Nevertheless, we found evidence of avoidance
development (e.g., large individuals require        effects in spruce trees due to negative
more resources and are more competitive) and/       association patterns with respect to the other
or by the different environmental conditions        two species. Several mechanisms may have
under which the recruitment of trees and            led to such a poor species interspersion. First,
saplings occurred (Espinosa et al. 2015). The       the spruce trees displayed a positive effect on
predominance of neutral diversity patterns in       the growth of their conspecific saplings (see
the ISAR analyses points to several possible        the next paragraph). Second, the spruce roots
explanations: existence of environmental            and litter deteriorate the edaphic conditions
filters, i.e. species responding to some kind of    under which the beech and fir saplings could
environmental heterogeneity (Espinosa et al.        grow sustainably, i.e. by water and base cation
2015); contrasting interactions that average        depletion, and raw humus accumulation from
each other out, i.e. null balance of positive and   and respectively, in the topsoil (Thelin et al.
negative interactions (Wiegand et al. 2007,         2002, Paluch & Gruba 2012, Paluch et al.
Punchi-Manage et al. 2015); demographic             2016). Third, despite being less shade-tolerant,
stochasticity due to local disturbance, like        spruce possesses some eco-physiological traits
those produced by herbivores, pathogens or          that translate in competitive advantages with
weather extremes (Wiegand et al. 2007, Gravel       respect to beech and fir: priority effects gained
et al. 2008). Various authors claimed that, in      by the early colonisation of canopy openings
accordance with the neutral theory (no species      (Jonášová & Prach 2004); best establishment
interactions), the ecological drift, dispersal      on thick humus layers and coarse woody
limitation and speciation alone can explain the     debris (Szewczyk & Szwagrzyk 1996, Orman
maintenance of tree diversity at local scales,      & Szewczyk 2015); highest rates of height
independent of species functional traits (Chave     growth (Stăncioiu & O’Hara 2006).
et al. 2002, Hubbell 2006).                            Only neutral neighbouring effects of trees on
   Signals of positive and negative interspecific   the vertical growth of heterospecific saplings
association were found in tree-sized beech and      were detected in the study stand. On the
spruce, respectively. The relatively high values    contrary, significant interactions were revealed
of mingling index observed in overstorey            between saplings and their nearest parent
beech trees is very likely linked to the status     trees. Like other studies performed in similar
of diversity accumulator revealed in beech          mixed forests from central Europe (Bosela
saplings. Niche complementarity in root             et al. 2015, Mina et al. 2018), we observed
foraging, shade-tolerance, leaf persistence and     a suppression in vertical growth of beech
crown growth architecture may also explain          saplings in the neighbourhood of conspecific
the positive association between beech and          trees. Many studies demonstrated the low
the two conifer species in the forest canopy.       self-tolerance of beech (being a strong self-

24
Gafta et al.                                         Neighbourhood-based evidence of tree diversity ...

competitor for both above-ground and below-        been induced by habitat microheterogeneity
ground resources) and its severe intraspecific     and/or small-scale disturbance, if the latter two
asymmetric competition due to high lateral         factors had acted at the same spatial scales. The
expansion (Pretzsch 2014, Pretzsch & Schutze       ecological response of tree species revealed
2016). Besides, beech performs better when         in this study are only valid in the range of
growing in mixtures with conifers thanks to        submontane-lower montane belt (800-1200
reduced intraspecific competition (Bosela          m) where the climatic conditions are optimal
et al. 2015, Pretzsch et al. 2010, Mina et al.     for the growth of beech and fir. Therefore, our
2018). On the other side, the unexpected           findings should be further validated through
positive effect (instead of interference) of the   replicated studies in similar old-growth, (sub)
nearest spruce trees on conspecific saplings       montane mixed forests.
could only arise by the greater benefit from          Both as juvenile and adult, the beech seems
parent nursing than the energetic loss due to      to be the main player in the equalising and
the asymmetric intraspecific competition. A        stabilising mechanisms of coexistence and
possible explanation for this self-favouring       diversity maintenance in the study forest stand.
process could be a facilitative below-ground       The dominant role of beech in mixture with fir
mechanism, like mutualist soil organisms           and spruce is also sustained by the unbalanced
(e.g., mycorrhizae) that enhance the growth of     pair-wise competitive interactions, since the
saplings (Das et al. 2008, Bennett et al. 2017).   negative effects of beech size-symmetric
As mentioned above, such a process may be          competition on the growth of fir and spruce
also responsible for the observed negative         is much stronger than the effect of fir and
spatial association between spruce and the         spruce size-symmetric competition on beech
other two species in the tree-size class. The      growth (Mina et al. 2018). Conversely, the
detected auto-facilitation is not necessary        low intermingling of spruce in the overstorey
in contradiction with the findings reported        leads to lower local tree diversity. This poor
from the same forest reserve by Duduman et         interspersion is supported by the segregated
al. (2010), who revealed an opposite effect        regeneration niche of spruce (e.g., on
(interference) on spruce saplings but in           deadwood) relative to beech and fir (Šebková
terms of their radial growth and considering       et al. 2012, Orman & Szewczyk 2015). The fir
neighbouring trees of any species. A similar       seems to occupy an intermediate rank between
example of auto-facilitation was documented        beech and spruce in terms of its contribution
in hemlock trees (Tsuga canadensis) growing        to maintaining the local tree diversity, but
in mixed coniferous-deciduous forests in           this outcome might be partly influenced
eastern North America (Woods 1984, Catovsky        by the low numeric share of fir, given the
& Bazzaz 2002).                                    strong dependency of association strength on
                                                   abundance (Vázquez et al. 2007). Based on
Limitations and concluding remarks                 similar processes observed in many similar
                                                   forest stands from western Carpathians (Paluch
Our findings are circumscribed to a (relatively    et al. 2019), we suppose that strong reduction
small) extent of 0.24 ha and the distance range    in stand density due to intensive exploitations
of 20 m (i.e., half of the smallest side of the    prior to reserve foundation has promoted the
plot), which might have prevented the detection    establishment of spruce regeneration (even at
of spatial patterns beyond that scale, namely      the lower limit of its altitudinal range) and has
signals of interactions between the large          decreased the recruitment of fir saplings.
trees. Also, the spatial-related patterns that
we attributed to biotic interactions could have

                                                                                                    25
Ann. For. Res. 64(1): 13-30, 2021                                                                          Research article

References                                                       Bosela M., Tobin B., Seben V., Petras R., Larocque G.R.,
                                                                   2015. Different mixtures of Norway spruce, silver fir,
                                                                   and European beech modify competitive interactions
Akashi N., 1996. The spatial pattern and canopy-                   in central European mature mixed forests. Canadian
  understory association of trees in a cool temperate,             Journal of Forest Research 45: 1577-1586. https://doi.
  mixed forest in western Japan. Ecological Research 11:           org/10.1139/cjfr-2015-0219
  311-319. https://doi.org/10.1007/BF02347788                    Bulleri F., Bruno J., Silliman B.R., Stachowicz J.J., 2016.
Arii K., Lechowicz M.J., 2002. The influence of overstory          Facilitation and the niche: implications for coexistence,
  trees and abiotic factors on the sapling community in an         range shifts, and ecosystem functioning. Functional
  old-growth Fagus-Acer forest. Ecoscience 9: 386-396.             Ecology 30: 70-78. https://doi.org/10.1111/1365-
  https://doi.org/10.1080/11956860.2002.11682726                   2435.12528
Baddeley A., Diggle P.J., Hardegen A., Lawrence T.,              Cammarano M., 2011. Co-dominance and succession in
  Milne R.K., Nair G., 2014. On tests of spatial pattern           forest dynamics: the role of interspecific differences in
  based on simulation envelopes. Ecological Monographs             crown transmissivity. Journal of Theoretical Biology
  84: 477-489. https://doi.org/10.1890/13-2042.1                   285: 46-57. https://doi.org/10.1016/j.jtbi.2011.06.031
Baddeley A., Rubak E., Turner R., 2015. Spatial point            Canham C.D., 1988. Growth and canopy architecture of
  patterns: methodology and applications with R.                   shade-tolerant trees: the response of Acer saccharum
  Chapman and Hall/CRC Press, Boca Raton.                          and Fagus grandifolia to canopy gaps. Ecology 69:
Baddeley A., Rubak E., Turner R., 2020. Spatial point              786-795. https://doi.org/10.2307/1941027
  pattern analysis, model-fitting, simulation, tests. R          Canham C.D., Le Page P.T., Dave C.K., 2004. A
  package ‘spatstat’ v1.64-1. https://cran.r-project.org/          neighborhood analysis of canopy tree competition:
  web/packages/spatstat/spatstat.pdf                               effects of shading versus crowding. Canadian Journal of
Bai X., Queenborough S.A., Wang X., Zhang J., Li B.,               Forest Research 34: 778-787. https://doi.org/10.1139/
  Yuan Z., Xing D., Lin F., Ye J., Hao Z., 2012. Effects           x03-232
  of local biotic neighbors and habitat heterogeneity            Catovsky S., Bazzaz F.A., 2002. Feedbacks between
  on tree and shrub seedling survival in an old-growth             canopy composition and seedling regeneration in
  temperate forest. Oecologia 170: 755-765. https://doi.           mixed conifer broad-leaved forests. Oikos 98: 403-420.
  org/10.1007/s00442-012-2348-2                                    https://doi.org/10.1034/j.1600-0706.2002.980305.x
Bândiu C., 1977. Lumina ca factor stabilizator al                Cavard X., Bergeron Y., Chen H.Y.H., Pare D., Laganiere
  compoziției ecosistemelor de amestec de brad cu fag.             J., Brassard B., 2011. Competition and facilitation
  In: Preda V. (ed.), Pădurea și spațiile verzi în actualitate     between tree species change with stand development.
  și perspectivă. Academia RSR, Cluj-Napoca, pp. 73-82.            Oikos 120: 1683-1695. https://doi.org/10.1111/j.1600-
Becker M., Drapier J., 1984. Rôle de l’allélopathie dans les       0706.2011.19294.x
  difficultés de régénération du sapin (Abies alba Mill.). I     Cenușă R., Popa C., Teodosiu M., 2002. Cercetări privind
  Propriétés phytotoxiques des hydrosolubles d’aiguilles           relaţia structură-funcţie şi evoluţia ecosistemelor
  de sapin. Acta Oecologica 5: 347-356.                            forestiere naturale din nordul ţării. Analele ICAS 45:
Becker M., Drapier J., 1985. Rôle de l’allélopathie dans           9-19.
  les difficultés de régénération du sapin (Abies alba           Chave J., Muller-Landau H.C., Levin S.A., 2002.
  Mill.). II Étude des lessivats naturels de feuillage, de         Comparing classical community models: theoretical
  litière et d’humus. Acta Oecologica 6: 31-40.                    consequences for patterns of diversity. American
Bennett J.A., Maherali H., Reinhart K.O., Lekberg Y.,              Naturalist 159: 1-23. https://doi.org/10.1086/324112
  Hart M.M., Klironomos J., 2017. Plant-soil feedbacks           Chen L., Mi X.C., Comita L.S., Zhang L.W., Ren H.B., Ma
  and mycorrhizal type influence temperate forest                  K.P., 2010. Community-level consequences of density
  population dynamics. Science 355: 181-184. https://doi.          dependence and habitat heterogeneity in a subtropical
  org/10.1126/science.aai8212                                      broad-leaved forest. Ecology Letters 13: 695-704.
Blundell A.G., Peart D.R., 2004. Density-dependent                 https://doi.org/10.1111/j.1461-0248.2010.01468.x
  population dynamics of a dominant rain forest canopy           Chesson P., 2000. Mechanisms of maintenance of species
  tree. Ecology 85: 704-715. https://doi.org/10.1890/01-           diversity. Annual Review of Ecology and Systematics
  4101                                                             31: 343-366.
Bolker B.M., Brooks M.E., Clark C.J., Geange S.W.,               Chi X., Tang Z., Xie Z., Guo Q., Zhang M., Ge J., Xiong
  Poulsen J.R., Stevens M.H.H., White J.S.S., 2009.                G., Fang J., 2015. Effects of size, neighbors, and site
  Generalized linear mixed models: a practical guide for           condition on tree growth in a subtropical evergreen and
  ecology and evolution. Trends in Ecology and Evolution           deciduous broad­leaved mixed forest, China. Ecology
  24: 127-135. https://doi.org/10.1016/j.tree.2008.10.008          and Evolution 5: 5149-5161. https://doi.org/10.1002/
Bolte A., Kampf F., Hilbrig L., 2013. Space sequestration          ece3.1665
  below ground in old-growth spruce-beech forests –              Comita L.S., Muller-Landau H.C., Aguilar S., Hubbell S.P.,
  signs for facilitation? Frontiers in Plant Science 4: 322.       2010. Asymmetric density dependence shapes species
  https://doi.org/10.3389/fpls.2013.00322                          abundances in a tropical tree community. Science 329:

26
Gafta et al.                                                       Neighbourhood-based evidence of tree diversity ...

   330-332. https://doi.org/10.1126/science.1190772                 org/10.1007/s40725-016-0031-2
Comita L.S., Queenborough S.A., Murphy S.J., Eck                Fox J.F., 1977. Alternation and coexistence of tree
   J.L., Xu K.Y., Krishnadas M., Beckman N., Zhu                    species. American Naturalist 111: 69-89. https://doi.
   Y., 2014. Testing predictions of the Janzen-Connell              org/10.1086/283138
   hypothesis: a meta-analysis of experimental evidence         Getzin S., Wiegand T., Wiegand K., He F.L., 2008.
   for distance- and density-dependent seed and seedling            Heterogeneity influences spatial patterns and
   survival. Journal of Ecology 102: 845-856. https://doi.          demographics in forest stands. Journal of Ecology
   org/10.1111/1365-2745.12232                                      96:        807-820.        https://doi.org/10.1111/j.1365-
Das A.J., Battles J.J., van Mantgem P.J., Stephenson                2745.2008.01377.x
   N.L., 2008. Spatial elements of mortality risk in old-       Gilbert B., Lechowicz M.J., 2004. Neutrality, niches, and
   growth forests. Ecology 89: 1744-1756. https://doi.              dispersal in a temperate forest understory. Proceedings
   org/10.1890/07-0524.1                                            of the National Academy of Sciences USA 101: 7651-
de la Cruz M., 2020. Individual diversity-area relationships.       7656. https://doi.org/10.1073/pnas.0400814101
   R package ‘idar’ v1.1. https://cran.r-project.org/web/       Gravel D., Beaudet M., Messier C., 2008. Partitioning
   packages/idar/idar.pdf                                           the factors of spatial variation in regeneration density
Diaci J., Rozenbergar D., Anic I., Mikac S., Saniga M.,             of shade-tolerant tree species. Ecology 89: 2879-2888.
   Kucbel S., Visnjic C., Ballian D., 2011. Structural              https://doi.org/10.1890/07-1596.1
   dynamics and synchronous silver fir decline in               Harms K.E., Condit R., Hubbell S.P., Foster R.B., 2001.
   mixed old-growth mountain forests in Eastern and                 Habitat associations of trees and shrubs in a 50-ha
   Southeastern Europe. Forestry 84: 479-491. https://doi.          neotropical forest plot. Journal of Ecology 89: 947-959.
   org/10.1093/forestry/cpr030                                      https://doi.org/10.1111/j.1365-2745.2001.00615.x
Diaci J., Rozenbergar D., Boncina A., 2010. Stand dynamics      Heiri C., Wolf A., Rohreb L., Bugmann H., 2009. Forty
   of Dinaric old-growth forest in Slovenia: are indirect           years of natural dynamics in Swiss beech forests:
   human influences relevant? Plant Biosystems 144: 194-            structure, composition, and the influence of former
   201. https://doi.org/10.1080/11263500903560785                   management. Ecological Applications 19: 1920-1934.
Diggle P.J., 2010. Nonparametric methods. In: Gelfand               https://doi.org/10.1890/08-0516.1
   A.E., Diggle P.J., Fuentes M., Guttorp P. (eds.),            Hiura T., Fujiwara K., 1999. Density-dependence and
   Handbook of spatial statistics. CRC Press, Boca Raton,           co-existence of conifer and broad-leaved trees in a
   pp. 299-316.                                                     Japanese northern mixed forest. Journal of Vegetation
Dobrowolska D, 1998. Structure of silver fir (Abies alba            Science 10: 843-850. https://doi.org/10.2307/3237309
   Mill.) natural regeneration in the ‘Jata’ reserve in         Hou J., Mi X., Liu C., Ma K., 2004. Spatial patterns and
   Poland. Forest Ecology and Management 110: 237-247.              associations in a Quercus-Betula forest in northern
   https://doi.org/10.1016/S0378-1127(98)00286-2                    China. Journal of Vegetation Science 15: 407-414.
Dobrowolska D., Veblen T.T., 2008. Treefall-gap structure           https://doi.org/10.1111/j.1654-1103.2004.tb02278.x
   and regeneration in mixed Abies alba stands in central       Hubbell S.P., 2006. Neutral theory and the evolution of
   Poland. Forest Ecology and Management 255: 34-69.                ecological equivalence. Ecology 87: 1387-1398. https://
   https://doi.org/10.1016/j.foreco.2008.02.025                     doi.org/10.1890/0012-9658(2006)87[1387:NTATEO]2
Doniță N., Chiriță C., Stănescu V. (eds.), 1990. Tipuri de          .0.CO;2
   ecosisteme forestiere din România. Centrul de Material       Hurtt G.C., Pacala S.W., 1995. The consequences of
   Didactic și Propagandă Agricolă, București.                      recruitment limitation: reconciling chance, history
Duduman G., Roibu C.C., Duduman M.L., Miron-                        and competitive differences between plants. Journal of
   Onciul M., 2010. The influence of competition and                Theoretical Biology 176: 1-12. https://doi.org/10.1006/
   dimensional-spatial characteristics of trees on their            jtbi.1995.0170
   radial growth in old-growth Slătioara forest, Romania.       Illian J.B., Penttinen A., Stoyan H., Stoyan D., 2008.
   AES Bioflux 2: 215-230.                                          Statistical analysis and modelling of spatial point
Duduman G., Tomescu C., Drăgoi M., Palaghianu                       patterns. John Wiley & Sons, Chichester.
   C., 2014. Variabilitatea dimensională a arborilor şi         Janík D., Adam D., Hort L., Král K., Šamonil P., Unar P.,
   diversitatea florei vasculare în amestecuri de răşinoase         Vrška T., 2014. Tree spatial patterns of Abies alba and
   cu fag din rezervaţia Codrul secular Slătioara. Bucovina         Fagus sylvatica in the Western Carpathians over 30
   Forestieră 14: 135-147.                                          years. European Journal of Forest Research 133: 1015-
Espinosa C., Cruz M., Jara-Guerrero A., Gusmán E.,                  1028. https://doi.org/10.1007/s10342-014-0819-1
   Escudero A., 2015. The effects of individual tree            Janík D., Král K., Adam D., Hort L., Šamonil P., Unar
   species on species diversity in a tropical dry forest            P., Vrška T., McMahon S., 2016. Tree spatial patterns
   change throughout ontogeny. Ecography 39: 329-337.               of Fagus sylvatica expansion over 37 years. Forest
   https://doi.org/10.1111/ecog.01328                               Ecology and Management 375: 134-145. https://doi.
Forrester D.I., Bauhus J., 2016. A review of processes              org/10.1016/j.foreco.2016.05.017
   behind diversity-productivity relationships in forests.      Johnson D.J., Bourg N.A., Howe R., McShea W.J., Wolf A.,
   Current Forestry Reports 2: 45-61. https://doi.                  Clay K., 2014. Conspecific negative density-dependent
                                                                                                                          27
Ann. For. Res. 64(1): 13-30, 2021                                                                        Research article

  mortality and the structure of temperate forests. Ecology      Assessing the scale-specific importance of niches and
  95: 2493-2503. https://doi.org/10.1890/13-2098.1               other spatial processes on beta diversity: a case study
Jonášová M., Prach K., 2004. Central-European mountain           from a temperate forest. Oecologia 159: 377-388.
  spruce (Picea abies (L.) Karst.) forests: regeneration         https://doi.org/10.1007/s00442-008-1214-8
  of tree species after a bark beetle outbreak. Ecological     Lasky J.R., Uriarte M., Boukili V.K., Chazdon R.L., 2014.
  Engineering 23: 15-27. https://doi.org/10.1016/j.              Trait­-mediated assembly processes predict successional
  ecoleng.2004.06.010                                            changes in community diversity of tropical forests.
Keren S., Motta R., Govedar Z., Lucic R., Medarevic M.,          Proceedings of the National Academy of Sciences
  Diaci J., 2014. Comparative structural dynamics of             USA 111: 5616-5621. https://doi.org/10.1073/
  the Janj mixed old-growth mountain forest in Bosnia            pnas.1319342111
  and Herzegovina: are conifers in a long-term decline?        Lieberman M., Lieberman D., 2007. Nearest-neighbor
  Forests 5: 1243-1266. https://doi.org/10.3390/f5061243         tree species combinations in tropical forest: the role of
King D.A., Wright S.J., Connell J.H., 2006. The                  chance, and some consequences of high diversity. Oikos
  contribution of interspecific variation in maximum tree        116: 377-386. https://doi.org/10.1111/j.2006.0030-
  height to tropical and temperate diversity. Journal of         1299.15370.x
  Tropical Ecology 22: 11-24. https://doi.org/10.1017/         Loosmore N.B., Ford E.D., 2006. Statistical inference
  S0266467405002774                                              using the G or K point pattern spatial statistics.
Klopcic M., Jerina K., Boncina A., 2010. Long-term               Ecology 87: 1925-1931. https://doi.org/10.1890/0012-
  changes of structure and tree species composition in           9658(2006)87[1925:SIUTGO]2.0.CO;2
  Dinaric uneven-aged forests: are red deer an important       Mina M., del Río M., Huber M.O., Thürig E., Rohner
  factor? European Journal of Forest Research 129: 277-          B., 2018. The symmetry of competitive interactions in
  288. https://doi.org/10.1007/s10342-009-0325-z                 mixed Norway spruce, silver fir and European beech
Kotanen P.M., 2007. Effects of fungal seed pathogens             forests. Journal of Vegetation Science 29: 775-787.
  under conspecific and heterospecific trees in a temperate      https://doi.org/10.1111/jvs.12664
  forest. Canadian Journal of Botany 85: 918-925. https://     Nagel T., Svoboda M., Diaci J., 2006. Regeneration
  doi.org/10.1139/B07-088                                        patterns after intermediate wind disturbance in an old-
Koukoulas S., Blackburn G.A., 2005. Spatial relationships        growth Fagus-Abies forest in southeastern Slovenia.
  between tree species and gap characteristics in broad-         Forest Ecology and Management 226: 268-278. https://
  leaved deciduous woodland. Journal of Vegetation               doi.org/10.1016/j.foreco.2006.01.039
  Science 16: 587-596.                                         Nagel T.A., Svoboda M., Rugani T., Diaci J., 2010. Gap
Kraft N., Adler P., Godoy O., James E., Fuller S., Levine        regeneration and replacement patterns in an old-growth
  J.M., 2015. Community assembly, coexistence, and the           Fagus-Abies forest of Bosnia-Herzegovina. Plant
  environmental filtering metaphor. Functional Ecology           Ecology 208: 307-318. https://doi.org/10.1007/s11258-
  29: 592­-599. https://doi.org/10.1111/1365-2435.12345          009-9707-z
Kuang X., Yuan Z., Lin F., Ye J., Wang X., Wang Y., Hao Z.,    Nakashizuka T., Kohyama T., 1995. The significance of
  2017. Conspecific density dependence and community             the asymmetric effect of crowding for coexistence in a
  structure: insights from 11 years of monitoring in an old-     mixed temperate forest. Journal of Vegetation Science
  growth temperate forest in Northeast China. Ecology            6: 509-516. https://doi.org/10.2307/3236349
  and Evolution 7: 5191-5200. https://doi.org/10.1002/         Orman O., Szewczyk J., 2015. European beech, silver
  ece3.3050                                                      fir, and Norway spruce differ in establishment, height
Kuninaga T., Hirayama K., Sakimoto M., 2015. Negative            growth, and mortality rates on coarse woody debris and
  canopy-understorey interaction shapes the sapling              forest floor - a study from a mixed beech forest in the
  bank of Fagus crenata in a cool-temperate, conifer-            Western Carpathians. Annals of Forest Science 72: 955-
  hardwood mixed forest. Plant Ecology 216: 1191-1202.           965. https://doi.org/10.1007/s13595-015-0492-7
  https://doi.org/10.1007/s11258-015-0501-9                    Pacala S.W., Levin S.A., 1997. Biologically generated
Kunstler G., Falster D., Coomes D.A., Hui F., Kooyman            spatial pattern and the coexistence of competing species.
  R.M., Laughlin D.C., Poorter L., Vanderwel M.,                 In: Tilman D., Kareiva P. (eds.), Spatial ecology: the
  Vieilledent G., Wright S.J., Aiba M., Baraloto C.,             role of space in population dynamics and interspecific
  Caspersen J., Cornelissen J.H.C., Gourlet-Fleury S.,           interactions. Princeton University Press, Princeton, pp.
  Hanewinkel M., Herault B., Kattge J., Kurokawa                 204-232.
  H., Onoda Y., Peñuelas J., Poorter H., Uriarte M.,           Paluch J., Bartkowicz L., Moser W.K., 2019. Interspecific
  Richardson S., Ruiz-Benito P., Sun I.-F., Ståhl G.,            effects between overstorey and regeneration in small-
  Swenson N.G., Thompson J., Westerlund B., Wirth C.,            scale mixtures of three late-successional species in
  Zavala M.A., Zeng H., Zimmerman J.K., Zimmermann               the Western Carpathians (southern Poland). European
  N.E., Westoby M., 2016. Plant functional traits have           Journal of Forest Research 138: 889-905. https://doi.
  globally consistent effects on competition. Nature 529:        org/10.1007/s10342-019-01209-y
  204­-207. https://doi.org/10.1038/nature16476                Paluch J., Gruba P., 2012. Effect of local species
Laliberté E., Paquette A., Legendre P., Bouchard A., 2009.       composition on topsoil properties in mixed stands

28
You can also read