ArticlePDF Available

A guild-based approach to assessing the influence of beech forest structure on bird communities

  • Reparto Carabinieri Biodiversità Castel di Sangro
A guild-based approach to assessing the influence of beech forest
structure on bird communities
Rosario Balestrieri
, Marco Basile
, Mario Posillico
, Tiziana Altea
, Bruno De Cinti
Giorgio Matteucci
Consiglio Nazionale delle Ricerche, Istituto di Biologia Agroambientale e Forestale, Montelibretti, RM, Italy
Corpo Forestale dello Stato, Ufficio Territoriale Biodiversità di Castel di Sangro-Centro Ricerche Ambienti Montani, Castel di Sangro, AQ, Italy
Consiglio Nazionale delle Ricerche, Istituto per i Sistemi Agricoli e Forestali del Mediterraneo, Rende, CS, Italy
article info
Article history:
Received 6 May 2015
Received in revised form 11 July 2015
Accepted 15 July 2015
Available online xxxx
ManFor C.BD.
Mixture models
Cavity nester
Forest management
The advantages of using birds as indicators of biodiversity and/or habitat quality in forest habitats rely
principally (1) on their relative ease of detection and (2) on their strong association with many forest fea-
tures. Here we investigate how forest structure can shape a bird community at the stand level (ca. 30 hec-
tares), using an expeditious sampling protocol and a simple but reasoned guild approach. We focused on
three beech (Fagus sylvatica) forests in Italy with different structures resulting from different manage-
ment histories: Cansiglio, Chiarano-Sparvera and Marchesale. We hypothesise that the abundance of
the selected guilds varies in relation to forest structure as a result of forest management; to test this
hypothesis we modelled guild abundance along a latitudinal transect in a similar habitat (i.e. beech for-
est) in order to offset the geographical and environmental sources of variability. Birds were surveyed
through replicated aural/visual point counts. Thus we identified four guilds: generalist cavity nesters
(TIT), generalist canopy nesters (WAR), insectivorous cavity nesters (INS) and granivorous canopy nesters
(FIN). Forest structure and deadwood were estimated for each of the 27 sampling points. Guild abun-
dance was estimated using the N-mixture models approach, which allows the mean abundance at each
sampling location and the detectability to be estimated. Both the mean abundance and detectability were
constrained to forest structure and time variables. Guild-estimated abundance was tested for differences
among sites using ANOVA. Mean guild abundance proved different only for two guilds. INS was more
abundant in Cansiglio than in Chiarano and Marchesale while between Chiarano and Marchesale there
was no significant difference. Also WAR abundance was greater in Chiarano, followed by Marchesale
and Cansiglio. Shelterwood management may result in a higher abundance of specialist guilds (i.e. insec-
tivorous cavity nesters), compared to other management options. Our results highlight this pattern in the
Cansiglio forest, where the age and the applied treatments resulted in a lower tree density and in a larger
mean diameter. Such features are often correlated with a larger amount of dead wood which, in turn,
promotes the presence of cavity nesters, otherwise rare-to-absent in the bird community. Forest struc-
ture is largely influenced by forest management that can be guided to benefit specific bird assemblages
by applying specific treatments focused at increasing structural diversity and at multi-functionality.
Ó2015 Elsevier B.V. All rights reserved.
1. Introduction
In the last few decades traditional forestry has begun to be
replaced in many western European countries by a more sustain-
able approach to harmonising forest harvesting with biological
conservation, carbon sequestration and long-term timber produc-
tion. A pan-European strategy has been set up and developed to
meet these goals (Rametsteiner and Mayer, 2004) and, regarding
biological conservation, great efforts have been placed on the use
and development of biodiversity indexes as well as indicator spe-
cies to provide guidance for forest management (Lindenmayer
et al., 2006, 2000; Noss, 1999).
Although the Forest Principles, drawn up during the United
Nations Conference on Environment and Development in 1992,
declared that the conservation of biological diversity was one of
the main goals in sustainable forest management, this purpose is
often difficult to pursue due to the uncertain response of the
animal community to forestry, which can be largely dependent
upon: (1) local conditions; (2) the taxa involved, and on (3) the
0378-1127/Ó2015 Elsevier B.V. All rights reserved.
Corresponding author.
E-mail address: (M. Basile).
Forest Ecology and Management xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Forest Ecology and Management
journal homepage:
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
time scale and spatial extent involved. For example, habitat loss is
widely recognised as one of the major causes of biodiversity
decline (Brooks et al., 2002; Pimm and Raven, 2000), but habitat
fragmentation could cause positive, negative or null effects on
biodiversity (Fahrig, 2003; Jansson and Andre, 2003; Villard
et al., 1999). Furthermore, the onset of edge effects after forest
cutting and their misinterpretation can lead to poor understanding
of the ecological system and of its reaction to disturbance (Craig,
2007; McCollin, 1998; Murcia, 1995; Ries et al., 2004). Hence,
biodiversity indexes which have, like all indices, a high synthesis
capability but limited power of analysis, should be used with great
caution and results carefully interpreted and transferred to other
forest contexts.
Another approach concerns the use of biological indicators,
defined as indicators of biodiversity and/or of abiotic conditions
as well as of changes in ecological processes (Spellerberg, 1994).
The fundamental principle of this approach is founded on the
assumption that the occurrence, or abundance, of an indicator spe-
cies could be informative about co-occurring species. Although this
is not necessarily true (Morrison, 1986), an indicator may provide
reliable information about species with similar characteristics, like
nesting sites or feeding habits (Severinghaus, 1981). This approach
is based upon the guild concept, which has been defined as a
‘group of species that exploit the same class of environmental
resources in a similar way’ (Root, 1967), and it may be a very pow-
erful tool in detecting changes in natural systems. The guild
approach, given the possibility to aggregate data, has proved useful
in data analyses and results interpretation, given that inventories
of forest birds may result in a large number of species and individ-
uals (Canterbury et al., 2000; Nikolov, 2009; Verner, 1984).
Birds have been widely studied as biological indicators in many
habitats such as wetlands (Croonquist and Brooks, 1991), streams
(Bryce et al., 2002), rangelands (Bradford et al., 1998) and forests
(Canterbury et al., 2000). The advantages of using birds as indica-
tors of biodiversity and/or habitat quality in forests rely principally
(1) on the relative ease of detection and (2) on their strong associ-
ation with many forest features. Abundance and compositions of
bird communities are affected by many forest features. These
include the dominance of coniferous or broadleaved trees
(Donald et al., 1998), the presence of plantation forests of
non-local species (Deconchat et al., 2009), the abundance of old
and dead trees, which usually provide plenty of nest sites
(Newton, 1994; Robles et al., 2011), silvicultural techniques (King
and DeGraaf, 2000), the spatial scale at which disturbances take
place (Drapeau et al., 2000), and the presence of primary cavity
nesters, i.e. woodpeckers, building microhabitats and nest cavities
(Martin and Eadie, 1999; Newton, 1994).
Most of the above studies focused on a medium/large spatial
extent and therefore refer to changes in the overall forest environ-
ment. Conversely, fewer studies have focused upon smaller
extents, investigating the influence of changes in forest structure
on the bird community at the level of a single forest stand (or smal-
ler) (Carrillo-Rubio et al., 2014; Nikolov, 2009; Spiering and Knight,
2005), which is the most common extent at which forest harvest-
ing takes place in Italy.
Here we investigate how forest structure can influence bird
community at the stand level (ca. 30 hectares), using an expedi-
tious sampling protocol and a simple but reasoned guild approach.
We hypothesise that the abundance of the selected guilds could
vary in relation to forest structure. To test this hypothesis we mod-
elled guild abundance along a latitudinal transect within three
beech forests (Fagus sylvatica) in order to offset the geographical
and environmental sources of variability. Furthermore, as forest
structure is strongly influenced by management, our study wanted
also to test how different forest management options may
influence bird communities.
2. Materials and methods
2.1. Study area
This research was developed within the sub-action
Forest Biodiversity (ForBD) of the Life+ project ManFor C.BD.
(Managing forests for multiple purposes: carbon, biodiversity and
socio-economic wellbeing). Among the seven study sites in Italy
we selected three beech stands of 30–35 ha located within (1)
Cansiglio Forest (Veneto Pre-Alps) in the Campo di Mezzo-Pian
Parrocchia State Reserve, (2) Chiarano-Sparvera Regional Forest
(Abruzzo, central Apennines) and (3) Marchesale State Reserve
(Calabria, southern Italy). Although all the study sites lie within
the Fagetum phytoclimatic band, they have a different structure
due to different management histories. The surface that was inter-
ested by the study is between 30 and 35 ha. The Cansiglio and
Marchesale stands are embedded within an extensive forested area
of mainly beech interspersed with a few scattered secondary pas-
tures, while the Chiarano-Sparvera stand represents the southern-
most portion of a large beech forest and is mostly surrounded by
The northernmost site is located within the Cansiglio Forest
(CA) (Veneto Region, 46°03
N, 12°23
E) lying on a limestone-marl
bedrock; it has a gently sloping morphology with small ridges
and valleys. Its elevation ranges from 1280 to ca. 1380 m a.s.l.;
mean yearly temperature is 5.6 °C and mean annual rainfall is
1660 mm. The stand is a mainly even-aged beech high-forest with
sporadic silver fir (Abies alba) and spruce (Picea abies) though
mature and ultra-mature high-forest and regeneration stages also
occur. The applied treatment is shelterwood, harvesting usually
occurring in small groups.
The Chiarano-Sparvera (CS) site is located in the middle of the
latitudinal transect within the central Apennines in Abruzzo
N, 13°57
E). The stand has an elongated shape and is sepa-
rated into two contiguous parts by a very narrow strip of meadows
and rocks. Elevation ranges from ca. 1700 to 1800 m a.s.l. and the
landscape morphology comprises an almost uniform (22–28.5°)
north-east facing slope on Cretaceous limestone. The mean yearly
temperature is 8.5 °C while average annual rainfall is 1100 mm.
The management type was coppicing with standards until the
early 1970s, when conversion to high-forest started with a first
thinning of sprouts from the stumps, leaving 1 or 2 stems per
stump maximum; the stand is currently a relatively young
even-aged high-forest.
The southernmost stand is located within the Marchesale State
Reserve (MA) in Calabria (38°30
N, 16°14
E). Its elevation ranges
from 1100 to 1200 m a.s.l. The landscape morphology consists of
an alternation of small hills (slope 40%), valleys and plateaux,
with a north-facing aspect. The bedrock is granite of the Serra
and Sila formation. Yearly mean temperature is 10.1 °C and mean
annual rainfall is 1880 mm. A small portion (5%) of mixed
beech-silver fir high-forest can be found at the border of the area.
It is an even-aged beech forest with both young and adult stages.
Treatment type is single tree selection.
2.2. Bird sampling protocol
Each stand was divided into nine equal sized (2.5–3 ha) patches,
in which centroid, birds were counted by means of unlimited
radius point counts (Blondel et al., 1981). The minimum distance
between sampling points was 130 m (mean ± sd = 164.8 ± 30.2).
Each of the 27 sampling points was surveyed 2–5 times
(mean = 3.5) in 2012, from mid-May to late-June. Within this per-
iod, sampling was replicated at intervals spanning from 1 day to
43 days (average = 8.5 days). We assume that surveys conducted
2R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
in such a short period ensure the closure assumption on the popu-
lations (Kéry, 2008). Surveys took place from sunrise to 11 a.m.,
and were always conducted by the same observer (RB) and lasted
/point; each detected bird was recorded at the species level by
both aural and visual cues, though the number of visual contacts
was negligible.
Given the small size of the patches and the proximity of sam-
pling points, to minimise dependency the full data-set was cen-
sored, discarding from the analysis forest-dwelling species with
large home ranges (e.g., woodpeckers Dendrocopos sp., blackbird
Turdus merula or mistle thrush Turdus viscivorus) as well as
non-forest species (e.g. the grey wagtail Motacilla cinerea) owls or
raptors (e.g. the common buzzard Buteo buteo). To test the spatial
independence among point counts, Moran’s I was calculated on
the censored data set using the mean count values for every
guild/site (Legendre and Legendre, 1998).
Guilds were first defined according to nesting (cavity vs. canopy
nesters) and feeding habits, according to Verner (1984). We identi-
fied four guilds: TIT = generalist cavity nesters (i.e. tits), WAR =
canopy generalist (i.e. mainly warblers), INS = insectivorous cavity
nesters (i.e. treecreeper and nuthatch), FIN = canopy granivorous
(i.e. finches, FIN) (Appendix S2). Though the FIN guild is comprises
taxa whose abundance generally depends more on seed productiv-
ity than on forest structure per se, we could not exclude a priori a
correlation between seed production and forest structure which,
in turn, could affect FIN abundance.
2.3. Forest structure
Forest structure and deadwood were sampled (clustered sam-
pling) for each of the 27 points within three circular plots (from
530 (CS) to 907 (MA, CA) m
/plot) placed at the vertices of a trian-
gle whose centre approximately overlapped with the centroid of
each patch; sampling methods followed the National Inventory
of Forest and Carbon (INFC, Tabacchi et al., 2006;
We analysed nine variables commonly used in forest-wildlife
studies: tree density (N, ha
), basal area (G, m
), mean
diameter of the dominant trees (ddom, cm), basal area mean
diameter (the diameter of the tree with the mean basal area, dg,
cm), mean height (hm, m), mean height of dominant trees
(hdom, m), stem volume (stemvol, m
), stem phytomass
(stemphyt, t ha
) and deadwood volume (deadwood, m
(Tabacchi et al., 2011). Forest variables were tested for differences
between sites using the Morisita index (Morisita, 1962),
implemented in the ANOSIM procedure (PAST; Hammer et al.,
2.4. Model building
Records of every species belonging to the same guild were
totalled and guild abundance was modelled separately, with each
guild as a single entity. Guild abundance models were built using
the N-mixture approach (Royle, 2004), having assessed the absence
of spatial dependence, except for the TIT guild in the Chiarano site
(Table 1). This N approach allows estimation of the state variable,
i.e., the mean abundance at each sampling location, and reveals
detectability, i.e. the probability of detecting a guild, should it be
present at a site. Our approach constrained the mean abundance
of guilds at sampling locations to forest structure, although a fixed
effect of forest identity was retained in the analyses, whereas
detectability was allowed to change among subsequent surveys:
abundance probability distribution ðb0þb1Nþb2Gþb3dg
þb4ddomþb5hm þb6hdom þb7stemvol
detectability binomial ðb0þb1survey1 þb2survey2 þb3
survey3 þb4survey4 þb5survey5Þ
Forest structure covariates were scaled before the analysis in
order to have a mean value of zero. For every guild, global models,
i.e. models in which all of the covariates were present, were tested
for goodness of fit (GoF), trying different distributions (Poisson (P),
Zero-Inflated Poisson (ZIP) or Negative Binomial (NB)) and upper
limits of integration (K) for the state variable. The GoF of the global
model was tested with the Pearson chi-square test, using a para-
metric bootstrap procedure (1000 re-samplings) to determine
whether the observed value is unusually large (MacKenzie and
Bailey, 2004).
2.5. Model selection
P-values and c-hat values were used to select the better model
to be used as global one. The global models of every guilds that
resulted in the best fit was chosen as a candidate for model build-
ing and selection procedure (Table 2). Each combination of forest
structure covariates were used to model the abundance, as well
as a null model without any covariates. The forest site fixed effect
was considered in each model, even in the null one. Models were
selected according to the Akaike’s Information Criterion (AIC)
(Akaike, 1973), taking into account that models with a score
variation >2 do not have the same empirical support, showing sub-
stantial differences (Burnham and Anderson, 2002). To be more
conservative, we considered only the models with a
AIC < 1 as
empirically supported. In addition, forest variables that showed a
Table 1
Spatial dependency of the average number of guild’s records calculated according to
Moran I for every site.
Cansiglio 0.168 0.012 0.118 0.030
Chiarano 0.487
0.006 0.105 0.140
Mongiana 0.132 0.235 0.152 0.186
p-value < 0.01.
Table 2
Models built using the N-mixture approach (Royle, 2004). Guild’s abbreviation are
shown in Table 1. NB = Negative Binomial; ZIP = Zero-Inflated Poisson; K = upper limit
of abundance integration.
Guild Global model parameters
and GoF values
Model selection
TIT Distribution = NB N 396.289 0
K= 30 N + G + stem.vol 396.819 0.530
p-value = 0.163 N + deadwood 396.856 0.567
c-hat = 1.21 Null model 397.251 0.962
INS Distribution = NB Ddom 265.376 0
K= 30 G + ddom 265.417 0.041
p-value = 0.229 hm 266.329 0.953
c-hat = 1.1 hdom 266.332 0.956
dg 266.353 0.977
ddom + deadwood 266.355 0.979
WAR Distribution = NB Null model 331.439 0
K= 30 hdom 331.846 0.407
p-value = 0.414 hm 331.847 0.408
c-hat = 1.02 Stem.vol 332.021 0.582
ddom 332.038 0.599
G 332.083 0.644
Dg 332.118 0.679
FIN Distribution = ZIP Null model 335.424 0
p-value = 0.252
c-hat = 1.09
R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx 3
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
Spearman rank correlation coefficient (
) > 0.80 were excluded
from the same model (Appendix: S1). Statistical analyses were
conducted using R and the packages ‘‘unmarked’’ (Fiske and
Chandler, 2011) and ‘‘AICmodavg’’ (Mazerolle, 2015).
The point abundance estimates of the guilds were tested for dif-
ferences among sites using parametric linear analysis of variance
(ANOVA), with the null hypothesis of no differences between point
abundance estimates for a given guild between sites. The Tukey
post hoc test was conducted. We calculated Shannon’s diversity
and evenness indexes (Maurer and McGill, 2011) based on guild
records to assess significant differences in site diversity and guild
3. Results
In all, we sampled 27 points in the three sites (9/site), collecting
1007 records of 44 species. Ten non-strict forest species and 12
forest species with large home ranges were excluded from data
analyses, resulting in a final inventory of 22 species and 769
records. The generalist cavity nesters guild (TIT) consisted of five
species, recorded 197 times, that of insectivorous cavity nesters
(INS) consisted of five species and 94 records, granivorous canopy
nesters (FIN) comprised four species and 275 records, and, finally,
generalist canopy nesters (WAR) consisted of eight species and 195
records (Appendix S2).
All comparisons of forest variables among sites proved
significant (Bonferroni corrected p-value < 0.01). Forest variables
that most contributed to differentiate the sites were tree density
(N), stem volume (stemvol) and stem phytomass (stemphyt),
representing a cumulative variability >94% in each paired
comparison (Table 3).
3.1. Model selection
The fixed site effect proved significant only for guilds INS and
WAR (p< 0.01). Estimated beta values for the abundance parame-
ters of the best model for each guild are given in Table 4. The best
model for the TIT guild was that with the state variable constrained
to tree density, although basal area, stem volume and dead wood
had a significant empirical support (Table 2). Best model estimates
indicate a negative response to the increase in density, while sec-
ondary model estimates indicate a positive response to basal
area/ha and a negative response to dead wood and stem volume.
Estimated guild abundance was 174.45 (CI
= 114–240) in CA,
132.62 (CI
= 81–193) in CS and 143.54 (CI
= 85–223) in MA
(Fig. 1).
The most supported model for the INS guild was that with the
state variable constrained to the mean diameter of the dominant
trees, although there was significant empirical support for the
models with response to basal area, mean height, mean height of
the dominant trees, mean diameter and dead wood. A slightly neg-
ative response is shown for dominant diameter in the best model,
whereas secondary models showed a negative response to mean
Table 3
Similarity percentage between the structural variables for the three forest sites. Means (±SD) are showed for each site and Bonferroni corrected probabilities for each pairwise
Variable Contribution % Cumulative contribution Mean (±SD) Morisita Anosim Bonferroni corrected p-value
for pairwise comparison
Cansiglio Chiarano Mongiana CAnCS CAnMO CSnMO
N 20.52 65.48 323 (39) 1370 (78) 510 (229) 0.37 1 0.63 1
Stemvol 5.50 83.03 543 (58) 289 (87) 497 (31) 0.61 1 1 0.71
Stemphyt 3.54 94.31 346 (37) 182 (56) 316 (20) 0.76 1 1 1
Dg 0.43 95.68 40.6 (1.8) 19.4 (4.6) 32.9 (1.7) 0.26 1 0.46 1
Ddom 0.39 96.91 49.2 (2.2) 30.9 (6.1) 46.1 (2.9) 0.73 1 1 1
Deadwood 0.36 98.06 18.1 (19.4) 15 (1.3) 5.55 (5.7) 0.17 0.67 1 0.36
Hm 0.25 98.84 26.6 (0.3) 14.4 (1.2) 23.4 (0.5) 0.05 0.35 0.52 1
Hdom 0.21 99.51 27.8 (0.2) 17.2 (1.1) 26.2 (0.6) 0.71 1 1 0.94
G 0.15 100 40.9 (3.9) 38.8 (5.4) 41.2 (3.4) 0.88 1 1 1
All variables <0.0001 0.0003 0.0009 0
Table 4
N-mixture model estimates of the bparameters of guild’s abundance.
Guild Parameter Mean SE
TIT Intercept (CA)
2.478 0.350
Fixed CS 0.930 0.690
Fixed MA 0.022 0.242
0.571 0.324
INS Intercept (CA)
6.198 1.931
Fixed CS
2.291 0.763
Fixed MA
1.268 0.395
0.072 0.038
WAR Intercept (CA)
2.088 0.279
Fixed CS
0.951 0.217
Fixed MA
0.696 0.241
FIN Intercept (CA)
1.624 0.197
Fixed CS 0.057 0.229
Fixed MA 0.270 0.248
p< 0.05.
p< 0.1.
p< 0.01.
Fig. 1. Guild’s abundance estimates for each site. Bars show the 95% confidence
4R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
height, dominant height and basal area mean diameter and a pos-
itive response to basal area and dead wood (Table 2). Estimated
guild abundance was 129.91 (81–186) in CA, 49.24 (19–85) in CS
and 49.05 (18–89) in MA (Fig. 1).
The best model for the WAR guild was the null model, although
models with mean height, mean height of the dominant trees, stem
volume/ha, mean diameter of the dominant trees, mean diameter
and basal area were highly supported as well (
AIC < 1)
(Table 2). Model estimates were positively correlated with these
variables. Estimated guild abundance was 72.63 (39–115) in CA,
188.01 (130–252) in CS and 145.70 (90–209) in MA (Fig. 1).
The most supported model for the FIN guild did not include
covariates. Hence the abundance of this guild did not relate to
the selected forest variables (Table 2). Estimated abundance was
45.64 (36–60) in CA, 43.09 (33–60) in CS and 34.85 (26–54) in
MA (Fig. 1).
Chiarano and Marchesale showed a similar point abundance
pattern, in respect to Cansiglio, where the community is repre-
sented by a higher proportion of cavity nester (Fig. 2).
Mean guild abundance differed only for INS and WAR, as
expected due to the significant influence of the site effect (Fig. 3).
INS abundance in Cansiglio was significantly greater than in
Chiarano and Marchesale (F= 32.03; df = 2; p< 0.01), while
between Chiarano and Marchesale there was no significant differ-
ence. WAR abundance was higher in Chiarano, followed by
Marchesale and Cansiglio (F= 127.1; df = 2; p< 0.01). FIN abun-
dance was also significantly different (F= 4.716; df = 2; p< 0.05),
but only between Cansiglio and Marchesale. Shannon diversity
and evenness indexes were similar among sites (Table 5), with
95% mostly overlapping confidence intervals.
4. Discussion
In this work, we show that forest structure, derived from a
specific management history, can affects birds community in dif-
ferent way. The existence of such relationships are well known,
but their quantification remain a large obstacle. Our approach, con-
sisting in an expeditious sampling protocol and in the use of recent
techniques to model bird abundance integrated with the guild
approach, demonstrate that many relationships could be actually
revealed. Another advantage is the applicability of such approach
at little spatial scale. Although point counts with unlimited radius
have been used, Moran test showed no significant spatial autocor-
relation between counts, with the exception of TIT guilds in
Chiarano. Moreover, our expert-based opinion (supported by many
tests carried out during other field work) is that in dense forests
birds vocalization are scarcely audible beyond 70 m. Actually, we
recorded the distance, but, as the 76% of the call were comprised
in a radius of 65 m, and the Moran test give us clear negative
results, we chose to retain every recorded call.
4.1. Influence of forest structure on guild abundance
Mature forests are usually highly suitable for cavity nesters or
bark feeding species (Bergner et al., 2015; Nikolov, 2009). Older
trees are more prone to branch fall, resulting in an increase in dead
wood and the subsequent onset of cavities (Peace, 1962).
Consequently, invertebrates could proliferate because of more
abundant suitable microhabitats, enhancing the food availability
for woodpeckers, that create even more cavities, and other
insectivores (Newton, 1994). Forestry often reduces the inverte-
brate biomass (Hill et al., 1990; Schowalter, 1990) generating
negative cascade effects on insectivorous birds (Czeszczewik
et al., 2014). Moreover, specialists are usually associated with older
forest stands and greater wood biomass (Caprio et al., 2008;
Donald et al., 1998). In mature forests large old trees play a
fundamental role in driving small-scale environmental
heterogeneity and generating those sets of characteristics that eli-
cit the spread of specialised cavity nesters (Kuuluvainen, 2002;
Newton, 1994). Our results show that the abundance of the INS
guild responds positively to an increase in basal area and dead
wood volume. Stands with a higher basal area and a lower tree
density could host larger and/or older trees which, in turn, could
result in a higher availability of nesting holes and suitable habitats
for invertebrate prey. INS occurs in the three sites with different
abundances, according to differences in tree diameters among
forests, resulting from both different management stages,
treatments or stand age. Deadwood explains the difference in INS
abundance only in conjunction with dominant diameter, as the
Chiarano site has a higher level of deadwood than Mongiana but
tree diameter is smaller than both at Cansiglio and Mongiana
forests (Table 3).
Fig. 2. Bird contacts collected at a given point (black) and mean point abundance estimates (grey).
R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx 5
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
Tits are usually the most abundant species among the sec-
ondary cavity nesters (Robles et al., 2011; Sage et al., 2011),
although they could be outnumbered in beech or primeval forests
by strictly insectivorous cavity nesters (Wesołowski et al., 2002;
Zangari et al., 2013). Overall TIT is the most abundant guild within
all sites and its estimated abundance is comparable across sites,
even though it shows a negative response to increasing tree den-
sity, stem volume and surprisingly to dead wood, while it
responded positively to basal area. We attribute this pattern to
the ability of this guild to outcompete other taxa as well as to its
adaptability, which results in a diffused dominance in managed
forests. Hence the abundance of TIT would decrease as the forest
became denser, while retaining some large trees could benefit it
through an increase in basal area distributed to larger trees.
Forest management could impact every species, although we
expect that generalists could be less affected (Carrillo-Rubio
et al., 2014). Not surprisingly, the null model for modelling WAR
abundance is the best supported (Table 2). Interestingly, WAR
abundance also shows a positive response to all of the covariates
related to tree size. Though this guild seems to benefit from an
increase in tree size, the large number of species contained in this
guild could level out the covariate effects, as the null model ranked
the lowest AIC. This guild has the most heterogeneous habits, with
more migrant species, some of which are also found in suburban
The model as also shows the lack of significant differences
among sites in the estimated abundance of the FIN guild, which
suggests no effect of forest structure on its abundance. We hypoth-
esise that their primary food (seeds) could be abundantly available,
with little if any relationship with forest management or treat-
ment. Moreover, the less marked territorial behaviour and with
respect to other species (Brichetti and Fracasso, 2013) could result
Fig. 3. Comparison (ANOVA) of mean estimated abundance of each guild between sites. Non-significant comparisons has a 0-crossing error bar.
6R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
in a more even distribution through each site. Finally, granivorous
birds may be negatively influenced by high canopy cover, prefer-
ring more open forest stands (Canterbury et al., 2000).
In Cansiglio and Mongiana tits represented by far the most
abundant portion of the community, and only at Chiarano were
warblers more abundant than tits. Conversely, finches are often
the smallest fraction of the community.
These results also show that the relationships with forest struc-
ture could not be ascribed to simple mechanisms beyond an inher-
ent ability to compete for resources. Even if the CI
(Table 5) are
largely overlapping, in Cansiglio the shelterwood treatment,
applied since long times, appears to result in the most equally
shared community according to evenness. The abundance of insec-
tivores suggests that the habitat is more suitable for obligated cav-
ity nesters as well as being rich in food. Indeed, in the insectivorous
guild there are many species which are more sensitive to habitat
alteration like the Eurasian treecreeper (Suorsa et al., 2004).
Shelterwood practice, with the retention of intact and older forest
patches could allow the persistence of some patches of the primary
habitat of these species.
At Chiarano the transitory high forest allows warblers to
increase, but the likely lower number of nesting cavities, because
of a younger stand age, does not allow the INS guild to further
increase. Being effective competitors among passerines, tits may
well exploit most of the available cavities found there. Chiarano
is the site with the lowest evenness index, which translates into
a community dominated by forest species also possibly adapted
to more open habitats.
In Mongiana the beech stand is older than at Chiarano and other
differences in stand structure are evident (Table 3). Despite this,
the guilds’ abundance patterns of two such sites are similar, prob-
ably thanks to their similar ages.
4.2. Implications for management and conservation
Forest harvesting is one of the main causes of forest habitat
alteration, with consequences such as the loss of fundamental
microhabitat for birds (Newton, 1994). Most of the previous stud-
ies have focused principally on the impact that forestry exerts on
bird populations or communities (e.g.,Suorsa et al., 2003; Bütler
et al., 2004; Gil-tena et al., 2008; Carrillo-Rubio et al., 2014). We
conjugated guild theory and abundance-based community models
into an easily-developed approach to study forest bird communi-
ties. We found that such an approach allows much of the informa-
tion contained in the original data set to be maintained, providing,
at the same time, an alternative or complementary approach with
respect to analyses focusing on single species. From a strictly
methodological point of view, a single-species approach could be
more useful to estimate abundance through spot-mapping, which
has a smaller sampling error and more power to detect changes
in abundance after forest harvest (Newell et al., 2013; Toms
et al., 2006). Point counts are used mostly for community-based
studies, where more species can be detected and less field work
is needed (Sutherland, 2006). Our approach could represent a com-
promise: we modelled each guild as a single entity, that responds
in a particular way to forest structure variation. Adopting this
approach allows inferences to be made about community shape
and relationships with forest structure both through observing
which the most abundant guilds are and through the relative rank-
ing among guilds. Beyond the dominant species approach (which is
usually affected by the abundance of generalist and less explana-
tory species like the common chaffinch, Fringilla coelebs), since
our method involves groups of species that exploit the same broad
resources, they can be more informative about the role of forest
variables with respect to the bird community.
Shelterwood treatments and the coexistence of different forest
management options at medium scale may result in a higher bird
species diversity compared to simpler, more impacting practices
like clear-cut or to more homogenous mature stands (King and
DeGraaf, 2000). Bird species richness is correlated with both larger
trees and sparser tree distribution (Carrillo-Rubio et al., 2014). Our
results support such pattern, as Cansiglio showed the highest diver-
sity (Table 5), with a lower tree density and a higher average diam-
eter. However, our approach is not limited to comparing diversity
which, due to its largely overlapping CI
, may level out actual dif-
ferences in bird communities. Forest structure is largely influenced
by forest management that can be guided to benefit specific bird
assemblages by applying specific treatments focused at increasing
structural diversity and at multi-functionality. Although the choice
of the assemblage is partly in the manager’s hands, we noted that,
among the INS guild, there are more protected and specialist spe-
cies. Moreover, we demonstrated that an easy-to-sample approach
may be as informative as a more complex and labour-intensive
approach like spot-mapping or mist-netting. Such an approach is
recommended whenever managers need valuable data in a short
This research was co-funded by the ManFor C.BD. Life+ project
(LIFE09 ENV/IT/000078). We are most grateful to A. Mancinelli who
greatly helped with songbird identification as well as many volun-
teers and students with helped in field work. C. Di Lena and A.
Daraio provided facilities at Cansiglio and Marchesale; the Forest
Service personnel at the Chiarano site invaluably helped with logis-
tics. We sincerely thank two anonymous referees who greatly
improved the manuscript.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at
Akaike, H., 1973. Information theory and an extension of the maximum likelihood
principle. In: Petrov, B.N., Cazakil, F. (Eds.), Second International Symposium
Information Theory. Akademiai Kidao, Budapest, Hungary, pp. 267–281.
Bergner, A., Avcı, M., Eryig
˘it, H., Jansson, N., Niklasson, M., Westerberg, L., Milberg,
P., 2015. Influences of forest type and habitat structure on bird assemblages of
oak (Quercus spp.) and pine (Pinus spp.) stands in southwestern Turkey. For.
Ecol. Manage. 336, 137–147.
Blondel, J., Ferry, C., Frochot, B., 1981. Point counts with unlimited distance. In:
Ralph, J.C., Scott, J.M. (Eds.), Estimating Numbers of Terrestrial Birds. Studies in
Avian Biology. Asilomar, California, USA, pp. 414–420.
Bradford, D.F., Franson, S.E., Neale, A.C., Heggem, D.T., Miller, G.R., Canterbury, G.E.,
1998. Bird species assemblages as indicators of biological integrity in great
basin rangeland. Environ. Monit. Assess. 49, 1–22.
Brichetti, P., Fracasso, G., 2013. Ornitologia Italiana. Sturnidae – Fringillidae, vol. 8.
Oasi Alberto Perdisa Editore, Bologna, Italy.
Brooks, T.M., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Rylands, A.B.,
Konstant, W.R., Flick, P., Pilgrim, J., Oldfield, S., Magin, G., Hilton-Taylor, C., 2002.
Habitat loss and extinction in the hotspots of biodiversity. Conserv. Biol. 16,
909–923, 10.1046/j.1523-1739.2002.00530.
Bryce, S.a., Hughes, R.M., Kaufmann, P.R., 2002. Development of a bird integrity
index: using bird assemblages as indicators of riparian condition. Environ.
Manage. 30, 294–310.
Table 5
Shannon diversity (H0) and Shannon evenness indexes for each study site. Indexes
have been calculated according to the guild’s records.
Site H
95% CI Evenness 95% CI
Cansiglio 1.314 1.281–1.352 0.930 0.900–0.967
Chiarano 1.255 1.277–1.353 0.877 0.897–0.968
Mongiana 1.293 1.258–1.363 0.911 0.879–0.977
R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx 7
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),
Burnham, K.P., Anderson, D.R., 2002. Model Selection and Multimodal Inference,
second ed. Springer-Verlag, New York.
Bütler, R., Angelstam, P., Ekelund, P., Schlaepfer, R., 2004. Dead wood threshold
values for the three-toed woodpecker presence in boreal and sub-Alpine forest.
Biol. Conserv. 119, 305–318.
Canterbury, G.E., Martin, T.E., Petit, D.R., Petit, L.J., Bradford, D.F., 2000. Bird
communities and habitat as ecological indicators of forest condition in regional
monitoring. Conserv. Biol. 14, 544–558.
Caprio, E., Ellena, I., Rolando, A., 2008. Assessing habitat/landscape predictors of bird
diversity in managed deciduous forests: a seasonal and guild-based approach.
Biodivers. Conserv. 18, 1287–1303.
Carrillo-Rubio, E., Kéry, M., Morreale, S.J., Sullivan, P.J., Gardner, B., Cooch, E.G.,
Lassoie, J.P., 2014. Use of multispecies occupancy models to evaluate the
response of bird communities to forest degradation associated with logging.
Conserv. Biol. 28, 1034–1044.
Craig, M.D., 2007. The short-term effects of edges created by forestry operations on
the bird community of the jarrah forest, south-western Australia. Austral. Ecol.
32, 386–396.
Croonquist, M.J., Brooks, R.P., 1991. Use of avian and mammalian guilds as
indicators of cumulative impacts in riparian-wetland areas. Environ. Manage.
15, 701–714.
Czeszczewik, D., Zub, K., Stanski, T., Sahel, M., Kapusta, A., Walankiewicz, W., 2014.
Effects of forest management on bird assemblages in the Bialowieza Forest,
Poland. iForest – Biogeosci. For.
Deconchat, M., Brockerhoff, E.G., Barbaro, L., 2009. Effects of surrounding landscape
composition on the conservation value of native and exotic habitats for native
forest birds. For. Ecol. Manage. 258, S196–S204.
Donald, P.F., Fuller, R.J., Evans, a.D., Gough, S.J., 1998. Effects of forest management
and grazing on breeding bird communities in plantations of broadleaved and
coniferous trees in western England. Biol. Conserv. 85, 183–197. http://
Drapeau, P., Leduc, A., Giroux, J.-F., Savard, J.-P.L., Bergeron, Y., Vickery, W.L., 2000.
Landscape-scale disturbance and changes in bird communities of boreal mixed-
wood forests. Ecol. Monogr. 70, 423–444.
Fahrig, L., 2003. Effects of habitat fragmentation on biodiversity. Annu. Rev. Ecol.
Evol. Syst. 34, 487–515.
Fiske, I.J., Chandler, R.B., 2011. Unmarked: an R package for fitting hierarchical
models of wildlife occurrence and abundance. J. Stat. Softw. 43, 1–23.
Gil-tena, A., Torras, O., Saura, S., 2008. Relationship between forest landscape
structure and avian species richness in NE Spain. Ardeola 55, 27–40.
Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. PAST: paleontological statistics
software package for education and data analysis. Palaeontol. Electron. 4, 1–9.
Hill, D., Roberts, P., Stork, N., 1990. Densities and biomass of invertebrates in stands
of rotationally managed coppice woodland. Biol. Conserv. 51, 167–176. http://
Jansson, G., Andre, H., 2003. Habitat composition and bird diversity in managed
boreal forests. Scand. J. For. Res. 18, 225–236.
Kéry, M., 2008. Estimating abundance from bird counts: binomial mixture models
uncover complex covariate relationships. Auk 125, 336–345.
King, D.I., DeGraaf, R.M., 2000. Bird species diversity and nesting success in mature,
clearcut and shelterwood forest in northern New Hampshire, USA. For. Ecol.
Manage. 129, 227–235.
Kuuluvainen, T., 2002. Natural variability of forests as a reference for restoring and
managing biological diversity in boreal Fennoscandia. Silva Fenn. 36, 97–125.
Legendre, P., Legendre, L., 1998. Numerical Ecology. Elsevier.
Lindenmayer, D.B., Franklin, J.F., Fischer, J., 2006. General management principles
and a checklist of strategies to guide forest biodiversity conservation. Biol.
Conserv. 131, 433–445.
Lindenmayer, D.B., Margules, C.R., Botkin, D.B., 2000. Indicators of biodiversity for
ecologically sustainable forest management. Conserv. Biol. 14, 941–950. http://
MacKenzie, D.I., Bailey, L.L., 2004. Assessing the fit of site-occupancy models. J.
Agric. Biol. Environ. Stat. 9, 300–318.
Martin, K., Eadie, J.M., 1999. Nest webs: a community-wide approach to the
management and conservation of cavity-nesting forest birds. For. Ecol. Manage.
115, 243–257.
Maurer, B.A., McGill, B.J., 2011. Measurement of species diversity. In: Magurran,
A.E., McGill, B.J. (Eds.), Biological Diversity. Frontiers in Measurement and
Assessment. Oxford University Press, Oxford, UK, pp. 55–65.
Mazerolle, M.J., 2015. Package ‘AICcmodavg’ 10.
McCollin, D., 1998. Forest edges and habitat selection in birds: a functional
approach. Ecography (Cop) 21, 247–260.
Morisita, M., 1962. Id-Index, a measure of dispersion of individuals. Res. Popul. Ecol.
(Kyoto) 4, 1–7.
Morrison, M.L., 1986. Bird population as indicators of environmental change. In:
Johnston, R.F. (Ed.), Current Ornithology. Springer, US, pp. 429–451.
Murcia, C., 1995. Edge effects in fragmented forests: implications for conservation.
Trends Ecol. Evol.
Newell, F.L., Sheehan, J., Wood, P.B., Rodewald, A.D., Buehler, D.A., Keyser, P.D.,
Larkin, J.L., Beachy, T.A., Bakermans, M.H., Boves, T.J., Evans, A., George, G.A.,
Mcdermott, M.E., Perkins, K.A., White, M.B., Wigley, T.B., 2013. Comparison of
point counts and territory mapping for detecting effects of forest management
on songbirds. J. Field Ornithol. 84, 270–286.
Newton, I., 1994. The role of nest sites in limiting the numbers of hole-nesting
birds: a review. Biol. Conserv. 70, 265–276.
Nikolov, S.C., 2009. Effect of stand age on bird communities in late-successional
Macedonian pine forests in Bulgaria. For. Ecol. Manage. 257, 580–587. http://
Noss, R.F., 1999. Assessing and monitoring forest biodiversity: a suggested
framework and indicators. For. Ecol. Manage. 115, 135–146.
Peace, T.R., 1962. Pathology of Trees and Shrubs. Clarendon Press, Oxford, UK.
Pimm, S.L., Raven, P., 2000. Extinction by numbers. Nature 403, 843–845. http://
Rametsteiner, E., Mayer, P., 2004. Sustainable forest management and pan:
European forest policy. Ecol. Bull. 51, 51–57.
Ries, L., Fletcher, R.J., Battin, J., Sisk, T.D., 2004. Ecological responses to
habitat edges: mechanisms, models, and variability explained. Annu. Rev.
Ecol. Evol. Syst. 35, 491–522.
Robles, H., Ciudad, C., Matthysen, E., 2011. Tree-cavity occurrence, cavity
occupation and reproductive performance of secondary cavity-nesting birds
in oak forests: the role of traditional management practices. For. Ecol. Manage.
261, 1428–1435.
Root, R.B., 1967. The niche exploitation pattern of the blue-gray gnatcatcher. Ecol.
Monogr. 37, 317–350.
Royle, J.A., 2004. N-mixture models for estimating population size from spatially
replicated counts. Biometrics 60, 108–115.
Sage, R.B., Hoodless, A.N., Hewson, C.M., Wilson, S., Clare, C. Le, Marchant, J.H.,
Draycott, R.A.H., Fuller, R.J., 2011. Assessing breeding success in common
woodland birds using a novel method. Bird Study 58, 409–420.
Schowalter, T.D., 1990. Invertebrate diversity in old-growth versus regenerating
forest canopies. Northwest Environ. J. 6, 403–404.
Severinghaus, W.D., 1981. Guild theory development as a mechanism for assessing
environmental impact. Environ. Manage.
Spellerberg, I.F., 1994. Monitoring Ecological Change, second ed. Cambridge
University Press, Cambridge, UK.
Spiering, D.J., Knight, R.L., 2005. Snag density and use by cavity-nesting birds in
managed stands of the Black Hills National Forest. For. Ecol. Manage. 214, 40–
Suorsa, P., Helle, H., Koivunen, V., Huhta, E., Nikula, A., Hakkarainen, H., 2004. Effects
of forest patch size on physiological stress and immunocompetence in an area-
sensitive passerine, the Eurasian treecreeper (Certhia familiaris): an experiment.
Proc. Biol. Sci. 271, 435–440.
Suorsa, P., Huhta, E., Nikula, A., Nikinmaa, M., Jäntti, A., Helle, H., Hakkarainen, H.,
2003. Forest management is associated with physiological stress in an old-
growth forest passerine. Proc. Biol. Sci. 270, 963–969.
Sutherland, W.J., 2006. Ecological Census Techniques. UK.
Tabacchi, G., Di Cosmo, L., Gasparini, P., Morelli, S., 2011. Stima del volume e della
fitomassa delle principali specie forestali italiane. Equazioni di previsione,
tavole del volume e tavole della fitomassa arborea epigea. Consiglio per la
Ricerca e la sperimentazione in Agricoltura, Unità di Ricerca per il Monitoraggio
e la Pianificazione Forestale., Trento, Italy.
Tabacchi, G., Scrinzi, G., Tosi, V., Floris, A., Paletto, A., Di Cosmo, L., Colle, G., 2006.
INFC Procedure di posizionamento e di rilievo degli attributi di terza fase (con
istruzioni per l’impiego degli applicativi NAV3 e RAS3). ISAFA/MIPAF – Corpo
Forestale dello Stato.
Toms, J.D., Schmiegelow, F.K.A., Hannon, S.J., Villard, M.-A., 2006. Are point counts of
boreal songbirds reliable proxies for more intensive abundance estimators? Auk
123, 438–454.
Verner, J., 1984. The guild concept applied to management of bird populations.
Environ. Manage. 8, 1–13.
Villard, M.-A., Trzcinski, M.K., Merriam, G., 1999. Fragmentation effects on forest
birds: relative influence of woodland cover and configuration on landscape
occupancy. Conserv. Biol. 13, 774–783.
Wesołowski, T., Tomiałojc, L., Mitrus, C., Rowinski, P., Czeszczewik, D., 2002. The
breeding bird community of a primaeval temperate forest (Białowie _
za National
Park, Poland) at the end of the 20th century. Acta Ornitol. 37, 27–45.
Zangari, L., Ferraguti, M., Luiselli, L., Battisti, C., Bologna, M.A., 2013. Comparing
patterns in abundance and diversity of hole-nesting birds in Mediterranean
habitats. Rev. Écol. (Terre Vie) 67, 1–8.
8R. Balestrieri et al. / Forest Ecology and Management xxx (2015) xxx–xxx
Please cite this article in press as: Balestrieri, R., et al. A guild-based approach to assessing the influence of beech forest structure on bird communities.
Forest Ecol. Manage. (2015),

Supplementary resource (1)

... Se sabe que las especies dependen de elementos estructurales de los bosques y que la perdida de ciertos componentes de los bosques maduros, como la alta cobertura de copas, los residuos leñosos y los árboles muertos en pie, afectan a las poblaciones de muchas especies que depende de estos elementos del hábitat (Newton 1994, Díaz et al. 2005, Czeszczewik et al. 2013, Balestrieri et al. 2015. Por ejemplo, en Suecia, árboles vivos grandes y maduros, árboles muertos en pie y residuos leñosos son recursos críticos para al menos 26% de los vertebrados e invertebrados en peligro, particularmente aves (Berg et al. 1994). ...
... Por ejemplo, para muchas especies de pájaros carpinteros proveen sitios para dormir y nidificar, además de alimento (Walankiewicz et al. 2002). Por otro lado, el incremento del área basal y el volumen de madera muerta se ha asociado positivamente con la abundancia de insectívoros en bosques de hayas (Abies alba y Picea abies) en Italia (Balestrieri et al. 2015). A su vez, sitios con área basal grande y baja densidad de árboles podrían presentar mayor disponibilidad de huecos para nidificar y mejores hábitats para predadores de invertebrados (Balestrieri et al. 2015). ...
... Por otro lado, el incremento del área basal y el volumen de madera muerta se ha asociado positivamente con la abundancia de insectívoros en bosques de hayas (Abies alba y Picea abies) en Italia (Balestrieri et al. 2015). A su vez, sitios con área basal grande y baja densidad de árboles podrían presentar mayor disponibilidad de huecos para nidificar y mejores hábitats para predadores de invertebrados (Balestrieri et al. 2015). Muhamad et al. (2013) encontraron que la riqueza de especies especialistas de bosque se asoció positivamente con la cobertura de copas en bosques montanos y tierras bajas dominados por Schima wallichii, Dysoxylum sp., y Sterculia coccinea (Indonesia). ...
El uso ganadero de los bosques de Nothofagus antarctica de Tierra del Fuego (Argentina) ha sido históricamente extensivo y poco planificado, y usualmente basado en la reducción de cobertura arbórea mediante cortas para aumentar la producción de forraje bajo su dosel. El objetivo de esta tesis fue analizar estructura y rasgos funcionales de las comunidades de aves terrestres, diurnas y residentes estivales (principalmente paseriformes), en diferentes estados del bosque de N. antarctica evaluando el efecto de variables ambientales a escala local y de paisaje, así como el potencial uso de las aves como indicadoras de impactos para mejorar las propuestas de estrategias de manejo foresto-ganadero existentes. Este trabajo se realizó en 4 Estancias (Los Cerros, Rolito, Las Hijas y Cabo San Pablo) ubicadas en el centro y este de la Isla Grande de Tierra del Fuego (Argentina). Se estudiaron 4 tipos de bosque: un tipo con raleos (Abiertos) y tres tipos sin raleos (Muy Cerrados, Cerrados y Muy Abiertos). Las aves se estudiaron en puntos de observación fijos, donde se realizaron conteos durante los meses de enero y febrero del 2017 y 2018 en Los Cerros, 2018 y 2019 en Rolito, y 2019 y 2020 en Las Hijas y Cabo San Pablo, para caracterizar estructura (composición, riqueza, densidad, biomasa e índices de diversidad) y rasgos funcionales (grupos tróficos, uso de sustratos, tipo de nidificación y estatus migratorio). En Los Cerros y Rolito también se analizó la susceptibilidad a la depredación de nidos artificiales. Los efectos analizados fueron: a escala local, la estructura forestal (altura dominante, área basal, cobertura de copas, densidad de árboles y diámetro medio), la cobertura del suelo (sotobosque, residuos leñosos, plantas no-vasculares, renovales y suelo sin vegetación), y la oferta alimenticia, vegetal (riqueza y cobertura total, de gramíneas y de dicotiledóneas consumidas por aves) y de la artropofauna (abundancia total y proporción de los órdenes más importantes); y a escala de paisaje, área, perímetro y forma del parche; área, número de parches, índice del parche más grande y conectividad de bosques y áreas abiertas; y total de bordes y densidad de bordes. Las variables fueron evaluadas mediante modelos lineales generalizados (GLM) y generalizados mixtos (GLMM), y multivariados (PCA, MRPP, IndVal, y CCA). Se pudo identificar mayor riqueza y diversidad de especies de aves en bosques Abiertos (raleados) y Muy Abiertos, dadas por la presencia de especies de áreas abiertas o de bordes de bosque. Por otro lado, se identificaron menores densidades de omnívoros y granívoras en bosques Cerrados y Muy Cerrados, respectivamente, vinculadas a los sustratos de alimentación. El uso de los sustratos difirió entre tipos de bosque, estando el uso de sustratos bajos y copas relacionado con su disponibilidad. Mientras que el uso de ramas y fustes, y la densidad de individuos volando se asociaron a la estructura forestal (cobertura de copas y área basal, entre otras). Por otro lado, la densidad de especies que nidifican en huecos y copa abierta, así como la densidad de residentes y migradoras, también difirieron entre tipos de bosque y se asociaron a la altura dominante y a la cobertura de copas. Los bosques raleados, si bien presentaron algunas diferencias con los Cerrados y Muy Cerrados (mayor riqueza de especies, diversidad, uso de sustratos bajos, densidad de aves volando y de migradoras), fueron parecidos en otros rasgos (ej. densidad de aves que utilizan ramas y copas o que nidifican en huecos), observándose en ellos incluso especies especialistas de bosque (Pygarrhichas albogularis y Aphrastura spinicauda). Si bien no se encontraron diferencias significativas en la susceptibilidad a la depredación de nidos artificiales entre tipos de bosque, los nidos que se encontraban en bosques Muy Abiertos fueron depredados más rápido. Este estudio permitió identificar especies indicadoras de bosques Cerrados y Muy Cerrados (A. spinicauda, P. albogularis), Abiertos (Spinus barbatus) y Muy Abiertos (ej. Tachycineta leucopyga). Por lo tanto, estas especies se podrían utilizar para el monitoreo de los raleos en estos bosques. Al analizar variables de diferentes escalas espaciales (local y paisaje), se pudo observar que las mismas influyeron sobre diferentes grupos de especies. Por ejemplo, la estructura forestal (área basal, cobertura de copas) y el área del parche de bosque tuvieron mayor efecto sobre P. albogularis. Se concluye que ciertas características a nivel local (ej. mayor área basal, cobertura de copas y de renovales, y proporción de himenópteros), así como a nivel de paisaje (ej. parches de bosque grandes e irregulares, alternancia con áreas abiertas), favorecen a una mayor diversidad de aves en los bosques de N. antarctica de Tierra del Fuego. El desarrollo de propuestas de manejo forestales y silvopastoriles que preserven o favorezcan la presencia de paisajes y parches de bosque con estas características contribuiría a que el manejo de estos bosques sea más sustentable.
... Similarly, unmanaged or protected forests are often characterized by old-growth attributes, including structures that need time to develop (Hedwall and Mikusiński, 2015;Paillet et al., 2015;Uotila et al., 2002). Hence, forest management that emphasizes timber production, by depleting such structures, can become one of the main drivers of the abundance and diversity of forest organisms (Balestrieri et al., 2015;Fedrowitz et al., 2014;Villard and Foppen, 2018). ...
... The abundance of species at a local scale in forests is largely dependent on the local forest structures (Balestrieri et al., 2015;Czeszczewik et al., 2015;Díaz et al., 2005). However, the surrounding landscape may influence the local abundance of the species, due to e.g. ...
... The composition of biological assemblages is known to differ between production forests and unmanaged forests, as well as between different management types (Balestrieri et al., 2015;Fischer and Lindenmayer, 2002;Paillet et al., 2010;Weibull and Ö stman, 2003). The species composition can vary independently of species richness or diversity, meaning that in homogeneous forest landscapes, high local diversity can be associated with low turnover of species among sites (Schall et al., 2018). ...
The variability in the amount and configuration of broad habitat types in the landscape, together with their structural complexity, influence observed biodiversity patterns. When considering structurally similar sites of the same habitat type, the variability in the abundance, species richness or diversity of organisms may be explained by the landscape context. To assess the numerical response of species to the landscape context, in terms of amount and configuration of forest environments, we investigated the bird assemblages of similarly structured forest habitats in an extensively managed forest region, encompassing different landscape contexts. We considered the numerical response of bird assemblages, in terms of abundance, species richness and diversity, and relative abundance of specific guilds, to the landscape context. We considered the forest cover at different spatial scales as a measure of habitat amount, while we quantified aspects of habitat configuration using various landscape metrics, and measured local forest structures. We found significant responses in multiple forest bird species to three important indices of forest structures: mean diameter of living trees, mean diameter of dead trees and volume of lying deadwood. Within similarly structured forest plots, bird assemblages showed responses linked with the landscape context, while plots with different habitat structure showed similar responses to the landscape context. In particular, there was a clear positive response of birds to the amount of broadleaf and mixed forest cover in the landscape. In addition, the distance between forest patches negatively affected species richness and diversity. Within landscapes, the increase of broadleaf in the existing forest area could boost abundance and diversity, decrease isolation levels for species dependent on broadleaves and enhance structural connectivity, generally favouring the majority of the species. Our findings suggest that the simple provision of habitat structures cannot represent a viable solution for biodiversity conservation and that the use of structural indicators of biodiversity like deadwood and age of canopy trees for assessing conservation value of forest needs to be integrated with landscape-scale indices. Our analysis clearly shows that the amount of habitat available in the surrounding landscape is linked with positive biodiversity responses. As human activities can alter both the provision of important habitat structures in stands across the landscape, as well as their overall landscape context, an integrated multi-scale biodiversity management is highly advisable.
... Insectivore densities were positively related to abundance of some insect groups (Hymenoptera, Diptera and Coleoptera, Appendix Table A5) documented as being prey for several bird species (Humphrey et al., 1970;Muñoz et al., 2017). Dominant height and basal area (more trees and/or larger trees) -the others explanatory variables associated with insectivore densitieshave also been identified as influential factors on the abundance of birds in Catalan Pyrenees (Ameztegui et al., 2018) and insectivores in Italy (Balestrieri et al., 2015). The influences of dominant height and basal area could be related to a greater stem surface in taller trees than in shorter trees, where insectivorous birds could explore for food (e.g., bark insects), although N. antarctica tree height is rather short even in the best quality sites (Lencinas et al., 2002), reaching just over 17 m. ...
Low intensity silviculture has been used to decrease the impact of forest harvesting, for example, on bird species and structural diversity. The objective of this work was to analyse the long-term effect of thinning on bird communities of Nothofagus antarctica forests in Tierra del Fuego (Argentina), compared with unthinned forests at two different locations. Thinning was performed 15 and 50 years ago at each location (ranches), therefore we also evaluated other common forest habitat types to differentiate these effects (location and time). We sampled four habitat types associated to overstory canopy cover (CC) categories: thinned (35–65% CC), and three unthinned forests (open with <35% CC, closed with 65–85% CC, and very closed with >85% CC), totalling 32 sampling sites (2 ranches × 4 canopy cover × 4 replicates). Bird assemblages’ structure and functional traits (e.g., richness, density, trophic groups, use of strata) were surveyed during two consecutive summers (2017–2020) at each site. We also characterized habitats by: (i) forest structure and ground cover (e.g., basal area, debris, and saplings); and (ii) food availability, considering understory plants consumed by birds (e.g., plant richness, grasses and dicots cover) and arthropods (e.g., total abundance). We evaluated the effect of CC, ranch, time, habitat and food availability by Generalised Linear Mixed Models and multivariate analyses (Multiple Response Permutation Procedure, Canonical Correspondence Analysis). In thinned forests, some bird structure and functional traits remained similar to closed forests; however, thinning increased bird species richness, being more similar to open forests. Effect of time could not be detected. CC and ranch were the factors that better described bird community structure, while forest structure, ground cover and food availability (e.g., dominant height, basal area, proportion of Hymenoptera) were the main drivers of most functional traits. The whole bird assemblage was better explained by 4–6 habitat structure and food availability variables depending on location (ranch). Results suggest thinning will benefit bird conservation if thinned forests maintain characteristics of mature forests (e.g., basal area > 40 m²/ha, shrub cover > 5%).
... Consequently, in human-dominated landscapes, such as managed forests, FD may decline following non-random species loss, i.e. loss of rare species with high contribution to FD. Indeed, management can have different effects on bird species, depending on their life history traits, resulting in some species being selectively more affected than others [22][23][24] . Bird assemblages of managed forests often occur at lower diversity, both numerical and functional, then in primeval forests, with management type or intensity often identified as significant factors determining numerical diversity 25,26 . ...
Full-text available
Functional diversity is linked with critical ecosystem functions, yet its relationship with numerical diversity, e.g. species richness, is not fully understood. The mechanisms linking changes of species richness, e.g. random and non-random species losses and gains, with changes of functional diversity become more relevant in the face of rapid environmental changes. In particular, non-random species changes including rare species may affect functional diversity, and the overall ecosystem function, disproportionately compared to random species changes including common species. In this study, I investigated how changes in numerical diversity of bird assemblages are related to functional diversity, and how the environment, and in particular forest management, influences such a relationship. I collected bird count data in the extensively-managed forest landscape of the Black Forest (Germany), at 82 sampling sites over three years. Data included species richness and abundance per site, and functional traits related to diet and habitat type for each species to compute functional diversity. By partitioning numerical diversity changes into five components using Price Equations, I calculated the contribution of random and non-random species losses and gains, and the abundance of common species, to functional diversity. Then I modelled these contributions as a function of several environmental variables describing broad forest conditions, and including forest management intensity. I found that, beside the major contribution of random species losses to functional diversity, non-random species losses also play a role, indicating that rare species that contribute more to functional diversity are often lost earlier than common species. The overall contribution to functional diversity of species losses is larger than that of species gains, pointing toward an ongoing simplification of the forest bird assemblage. Among all Price components, random species gains were influenced by management intensity, while other components were not influenced by any management variable. This highlight that potential conservation actions may not be effective in halting ecosystem functioning decline, as species gains do not result in increased functional diversity.
... Species richness and abundance increases from young cuts to mature stands, where stenotopic forest dwellers like Sitta europaea and woodpeckers concentrate. Also in beech forests (Balestrieri et al. 2015;Redolfi De Zan et al. 2016) tree diameter (> than 50 cm DBH), standing rotten trees, wood necromass and diversity of deadwood are essential resources for primary and secondary hole-nesting birds and shelterwood management seems the best solution. Old stands with beech trees of >30 cm and 26/32 m height with neighbouring old conifer stands are important for the black woodpecker in the alpine beech forest (Colpi et al. 2008). ...
Full-text available
We re-examined the knowledge on the relationships between fauna and forest management in Italy. We focused our attention on Vertebrates, including small mammals (Rodents and Soricomorphs), bats, birds, reptiles and amphibians, and on carabid and saproxylic beetles. We found that nearly all groups respond in accordance with Angelstam’s scheme of the traditional forestry cycle, with open habitat specialists followed by generalists and forest specialists when pristine conditions have been attained, though with some differences for some taxa. In all groups, some forest “superspecialists” or old growth indicators could be found, in (sub)Mediterranean forests some carabid beetles could be defined as “dendrophilous”, i.e. associated with the presence of old trees. Poplar plantations are poor substitutes for native forests, as well as exotic conifers We propose a model for an ideal forest structure that may efficiently preserve the faunal components, at the microscale level this coincides with the structure of an unmanaged old growth forest stand, marked also by a deep and humus-rich soil layer. Some caveats at the landscape scale have dealt with in relation to the size of the designed area, in order to avoid fragmentation effects. A sustainable harvesting scheme has indicated in the “systemic silviculture” proposed by Ciancio and other scientists.
... Commercial plantations are generally categorized as an unfavorable environment owing to the frequent disturbance caused by humans and the low availability of food sources (Aratrakorn et al., 2006). Foraging guilds often respond to environmental conditions differently (Balestrieri et al., 2015). A decline in the avian population of long-distance migrant species of insectivore birds has been demonstrated by Vickery et al. (2014) and Gregory et al. (2007) in commercial plantation zones. ...
Full-text available
Punjab is an agrarian state nurtured by the river Beas and Satluj with a maximum gross cropped area of 98.5% in India. River Beas had been designated as "Conservation Reserve" in 2017 and "Ramsar site" in 2019. Agriculture affects 87% of the globally threatened bird species. Birds are the most noticeable and specialized species in the river habitats; hence the abundance and distribution of birds are often readily interpreted in the context of river health and vice versa. The chapter has been designed to understand the impact of agriculture, urbanization, tree diversity and wetlands along the river on avian diversity based on literature available for the state of Punjab. The chapter aims to generate sustainable management strategies for the conservation of avian diversity at Beas river conservation reserve without hampering the development of the region. The present work reveals that agriculture intensification and urbanization are major concern for avian diversity conservation as both these factors negatively impacts the habitat specialist bird species and favors generalist and insectivore species. The study concludes that significant number of species recorded have specific niche area requirements that are completed by the river's sub-habitats including feeding, foraging, roosting and nesting therefore the whole area needs protection as a single unit.
Full-text available
In order to determine the relationships between bird assemblages and forest habitat, we conducted surveys for bird assemblages in different forest habitats in the Khentii Mountain region, Northern Mongolia. A total of 1730 individuals belonging to 71 species from 23 families of 11 orders were recorded. Our findings revealed that passeriformes are the most species-rich order, accounting for 86.2% of the total species. The dominant species were Anthus hodgsoni, Parus major, Poecile palustris, and Sitta europaea in study area. Non-metric multidimensional scaling (NMDS) and permutation multivariate analysis of variance (PERMANOVA) showed that bird assemblages were affected by forest habitat types. Our findings also showed significant relationships between bird assemblages and canopy height and ground cover vegetation structure, whereas there were no relationships between altitude and other habitat variables. Thus, maintaining diverse forest habitats or restoring forest would play a key role in bird conservation and sustainable management of forest areas.
Full-text available
Forests are undergoing through a slow but steady process of forest area change, which happen to be different across countries. Predictions of future scenarios highlight the role of agriculture in shaping land-use change in Europe, with possible local benefits for forests. Thus, the role of management in shaping the forests of the future will increase according to the World’s demand for timber, clean water, biodiversity, clean air and other ecosystem services. Research is testing and developing new management strategies to halt completely biodiversity loss. The threats faced by biodiversity in managed forests constitute the challenges of future forest ecological research. The response of birds to environmental changes is a consequence of the close-knit species-habitat relationships they evolved and are adapted to. Adaptations to the forest environment are structural, behavioural and physiological. The alteration of the forest structures and their spatial allocation across landscapes creates threats to bird populations that need to be taken into account in conservation-oriented forest management. The present thesis aims to assess the numerical response of the bird assemblage to forest structures and its relevance for management. The general aim is focused on close-to-nature forest management and on the retention of structural elements as a conservation tool. The numerical response of the bird assemblage is investigated in terms of species diversity and abundance. The general aim is achieved by focusing on the following research questions: i. Which forest and landscape characteristics best explain the abundance and diversity of forest birds? ii. Which management practices can be developed for the conservation of birds in managed forests? This thesis is carried out within the framework of the ConFoBi Research Training Group (Conservation of Forest Biodiversity in multiple-use landscape of Central Europe). ConFoBi assesses whether and how structural retention measures contribute to the conservation of forest biodiversity in multiple-use landscapes of the temperate zone. The research programme of the present thesis has been implemented in the southern Black Forest, Germany, as a model system for temperate forests. The Black Forest is a forest-dominated low mountain range within a multiple-use landscape typical of central Europe, where 135 quadratic 1-ha-plots where selected within the forested areas. Plots were selected along two major environmental gradients: forest vertical complexity, represented by the number of standing dead trees found in each plot; landscape composition, represented by the amount of forest cover within 25 km2 surrounding each plot. All plots were inventoried by measuring the diameter of every tree (above 7 cm diameter at breast height) and the amounts of laying and standing dead wood. Because habitat trees are an important structural element in retention approaches, for each study plot the 15 largest trees were mapped and their existing microhabitats (such as hollows, cracks, and cavities) quantified. Landscape patterns were described using a range of metrics, including amount of forest cover and landscape metrics. Bird data were collected by employing point counts performed at the centre of each plots and repeated two or three times per sampling season (2017-2018). Abundance and diversity of birds were analysed using hierarchical N-mixture models and generalized linear (mixed) models, respectively. The results highlight that current retention practices are well below the retention levels needed by the bird assemblage to have abundance and diversity similar to those of unharvested forests, assessing the levels around 40-60% of retained trees. Retention practices can, however, be improved by retaining trees of higher ecological importance. Trees with a diameter 20 cm larger than the surrounding trees, with cavities, rot holes or concavities can potentially fulfill the habitat requirments of a large array of species, including woodpeckers and the related cavity-users community. The presence of those trees across the landscape can ensure the continuity of resources, while favouring more natural tree species composition will benefit bird species diversity. The response of the entire bird assemblage to the configuration of forest patches indicates that in landscapes with less aggregated forest patches, the abundance of the bird assemblage can potentially be lower than in landscapes with closer forest patches. I conclude that the establishment and assemble of the bird community is a hierarchical spatial process. Small environmental elements determine the occurrence and observed abundance of populations. The location of such elements determines which species and how many individuals of each species are found across the landscape. Over the entire landscape, the amount of available habitat and its configuration determines how many species can be found at a given site. Therefore, I suggest a hierarchical set of actions should be considered, each benefiting birds at different levels of biological organization and spatial scale. Actions to preserve forest area and avoid fast land-use changes should have higher priority. The provisioning of large trees and deadwood would then ensure that the forests retain those structures and functions that support the establishment of bird populations in optimal habitats.
Because of its size, high segmentation within the global market and challenges driving its development, the separation technology sector could benefit from a bio-inspired approach to innovation to create more efficient and sustainable solutions. The potential for bio-inspired innovation is still largely untapped and Biologically-Inspired Design (BID) methods and tools are still largely underutilised, especially within the industry. A new BID method, called Guild-Based (GB) BID, is proposed to create a database of a large set of biological solutions – identified by a function – where biological information is structured to be more effective and usable within the industrial environment. A database for the separation technology sector has been set up and populated with 118 relevant biological solutions responding to the main function ‘to separate’. The database has been utilised to generate several clusters of solutions depending on the level of detail of the formulated problem. In particular, these include broad design principles of separation, taxonomies of biological solutions for specific separation problems and novel design concepts for two specific separation technologies (a desalination technology and an antibacterial surface). Furthermore, because of a large dataset of biological solutions, the possibility of determining the frequency of occurrence of specific separation strategies in nature can trigger reflections on the impact of existing separation technologies and taking decisions on future related R&D paths. More tests need to be conducted in the industrial environment; however, the results achieved so far indicate that the method proposed can indeed be instrumental to generating innovative ideas of interest to the separation technology sector.
Full-text available
In addition to providing an understanding of processes within a catchment system, isotopic techniques have been instrumental in providing reconstructions of catchment climate and other environmental indicators at various time scales. Many recent changes are a direct consequence of anthropogenic activities. Isotopic analysis serves as a valuable tool for distinguishing between natural variations in long-term climatic patterns and anthropogenic effects, yielding improved understanding of natural feedback mechanisms and the development of realistic remediation strategies. This chapter discusses the examples of isotopic techniques that have been applied to understand several types of ongoing and recent environmental changes, and in paleo-environmental studies. It discusses isotope geochemistry, hydrology, and climatology to look at new ways of applying isotopic tracing techniques to provide information on environmental change. It also gives an overview on how isotopic indicators are being applied in investigations of environmental change in continental settings.
This study investigated the occurrence of 25 forest bird species in relation to stand and landscape habitat composition in managed boreal forests in central Sweden. The number of species in 10 km transects was positively correlated to the proportions of forest > 40 yrs and Mixed forest >40 yrs, as well as the number of fragments of the latter, while negatively correlated to the proportions of clear-cuts and young forest. Transects with >60% older forest including > 6% mixed habitats showed the highest number of bird species and individuals. The positive effect on species numbers of mixed forest was stronger in the study area where deciduous rich habitats were generally less abundant. Bird species richness in points of similar habitat was negatively correlated to the degree of fragmentation of the surrounding older forest. However, in a species-by-species analysis at the transect level no effects of fragmentation were found. The number of bird species at points was positively influenced by the increase from 0 to 5% deciduous trees, while no effect of higher deciduous site proportions showed. The hazel grouse (Bonasa bonasia) appeared as a significant indicator of high bird diversity at the transect (landscape) level, by all methods used. However, the conclusion was reached that forests where the hazel grouse, blue tit (Parus caeruleus), treecreeper (Certia familiaris), jay (Garrulus glandarius) and capercaillie (Tetrao urogallus) are present are most likely to hold the potential for high bird species richness.
An age-old pursuit of biologists has been to relate the distribution and abundance of organisms to some aspect of their environment. Factors influencing their evolution, reproductive success, dispersal and migration, and other aspects of their ecology have been investigated. These pursuits have ranged from qualitative descriptions of plant and animal distributions and associations to quantitative statistical analyses of the interrelationships of plants, animals, and their environment.