Content uploaded by José María Fernández-García
Author content
All content in this area was uploaded by José María Fernández-García on Oct 19, 2023
Content may be subject to copyright.
Abstract
The cerambycid beetle Rosalia alpina is associated with temperate, broadleaved (mainly
beech) forests containing dead or decaying wood. This species is protected under the Habi-
tats Directive of the European Union. Given its narrow ecological niche and limited dispersal
abilities, habitat fragmentation is a conservation concern for populations of R. alpina. In
order to maximise the effectiveness of habitat restoration, a scientifically sound procedure
for patch selection is needed. In Gipuzkoa (N Spain), we used Light Detection and Ranging
(LiDAR) images to search for 20x20-m cells matching the parameterisation of a predictive
and local habitat model for R. alpina at the tree level. The cells selected under quantitative
criteria were clustered to identify potential habitat patches. Conefor Sensinode software
was used to estimate the importance of each of those patches in terms of the connectivity
Rosalia alpina connectivity
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
José María Fernández-García1, Javier Sesma2, Eugenio Moreno1, Valentín Mugarza3
Prioritizing forest patches to enhance habi-
tat restoration and connectivity for the
endangered saproxylic beetle Rosalia alpina
(Coleoptera, Cerambycidae): a modelling
approach.
Priorización de parches forestales con el fin de
favorecer la restauración del hábitat y la conecti-
vidad del escarabajo saproxílico Rosalia alpina
(Coleoptera, Cerambycidae): un enfoque de
modelización.
1Hazi Foundation
Granja Modelo, 01192 Arkaute, Spain
2Provincial Council of Álava
Plaza de la Provincia, 01001 Vitoria, Spain
3Provincial Council of Gipuzkoa
Plaza de Gipuzkoa, 20004 San Sebastián, Spain
Contact: jofernandez@hazi.eus
https://doi.org/10.21630/mcn.2023.71.09
of the population, based on dispersal distances of R. alpina and the probabilities of dis-
persing events among patches. We identified 380 potential habitat patches, mostly in the
south-eastern sector of the study area, which were classified into “core areas” and “con-
necting areas”. The performance of the model was tested in the field (61% of correct
assignations), although the actual occurrence of R. alpina within the habitat patches should
be assessed in the future. This model represents a step forward in guiding the cost-effective
implementation of conservation activities, through a strict preservation of the current core
habitat patches and an increase in the size of connecting patches. Therefore, we show
that connectivity models combining remote sensing data and local habitat selection can
be an aid in conservation planning and restoration actions, probably outperforming less
efficient strategies, such as random or expert selection.
Key words: Beetle, fragmentation, conservation, beech, Fagus sylvatica.
Resumen
El escarabajo cerambícido Rosalia alpina está asociado a los bosques templados de fron-
dosas (principalmente hayas) que albergan madera muerta o en descomposición. La espe-
cie está protegida bajo la Directiva de Hábitats de la Unión Europea. Dado su estrecho
nicho ecológico y su limitada capacidad de dispersión, la fragmentación de su hábitat es
un problema de conservación para las poblaciones de R. alpina. Para maximizar la efecti-
vidad de la restauración del hábitat, se necesita un procedimiento científicamente sólido
para la selección de parches forestales sobre los que intervenir. En Gipuzkoa (norte de Espa-
ña) utilizamos imágenes LiDAR para buscar celdas de 20x20 m que se ajustaran a la para-
metrización de un modelo predictivo local de hábitat para R. alpina a nivel de árbol. Las
celdas seleccionadas bajo criterios cuantitativos fueron agrupadas para identificar parches
de hábitat potencial. Se utilizó el software Conefor Sensinode para estimar la importancia
de cada uno de estos parches en términos de conectividad para la población, con base en
las distancias de dispersión de R. alpina y la probabilidad de eventos de dispersión entre
parches. Se identificaron 380 parches de hábitat potencial, mayoritariamente en el sector
sureste del área de estudio, que fueron clasificados como “áreas centrales” o “áreas de
conexión”. La ejecutoria del modelo fue testada en campo (61 % de asignaciones correc-
tas), si bien la presencia real de R. alpina en los parches de hábitat debería ser confirmada
en el futuro. Este modelo constituye un avance para guiar la implementación de actividades
de conservación de manera eficiente, mediante la protección estricta de “áreas centrales”
y el incremento del tamaño de las “áreas de conexión”. Por lo tanto, mostramos que los
modelos de conectividad que combinan datos de detección remota y selección local del
hábitat pueden asistir en la planificación de la conservación y en las acciones de restaura-
ción específicas, seguramente mejorando los resultados de estrategias menos eficientes,
como la selección al azar o basada en criterio experto.
Palabras clave: Escarabajo, fragmentación, conservación, haya, Fagus sylvatica.
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Laburpena
Rosalia alpina kakalardo cerambicidoa atxikita dago egur hila edota deskonposizioan duten
baso hostozabal epeletara (batik bat pagadiak). Espeziea Europar Batasuneko Habitat
Arteztarauak babesturik dago. Txoko ekologiko mugatua izateagatik eta dispertsiorako
ahalmen murritza, habitataren zatiketa kontserbazio arazo handia da R. alpina populazio-
entzat. Habitat berreskurapena ahalik eta eraginkorrena izan dadin beharrezkoak dira
zientzia prozedura sendoak, esku hartuko den baso orbain egokienak aukeratzeko. Gipuz-
koan (Espainia iparraldean) LiDAR irudiak erabili ditugu 20x20 m-ko zelden bidez R.alpi-
na-ren habitataren modelo prediktibo lokalera doitzeko, zuhaitz mailan. Irizpide
kuantitatiboen bidez aukeratutako zeldak multzoka bildu dira, habitat potentzialen orbai-
nak antzemateko. Conefor Sensinode softwarea erabili da orbain bakoitzaren garrantzia
estimatzeko populazioaren konektagarritasunaren aldetik, R. alpina-ren dispertsio distan-
tzian oinarrituz eta orbainen arteko dispertsio aukeren probabilitatea lortzeko. Habitat
potentzialeko 380 orbain identifikatu dira, gehienbat ikerketa arearen hego-ekialdeko sek-
torean, sailkaturik “gune zentralak” eta “konexio gune” eran. Exekuziorako modeloa ingu-
rune naturalean probatu da (%61 antzemate zuzen), baina R.alpina-ren presentzia erreala
habitateko orbainetan egiaztatu egin beharko litzateke etorkizunean. Modelo hau aurre-
rapausoa da kontserbazio ekintzak modu eraginkorrean garatzeko, “gune zentralen”
babes zorrotza bultzatuz eta “konexio guneen” tamaina areagotuz. Horrenbestez, era-
kusten dugu konektagarritasun urrutiko detekzioko datuak eta habitaten selekzio lokala
uztartzen duten modeloek errestaurazio espezifikoko ekintzetan eta kontserbazio plan-
gintzan lagundu dezaketela, ziurrenik eraginkortasun baxuagoko estrategien emaitza
hobetuz, kasurako, zoriz eginiko aukeraketa edota adituen iritzietan oinarrituta.
Gako hitzak: kakalardoa, zatiketa, kontserbazioa, pagoa, Fagus sylvatica.
Introduction
Connectivity is a landscape feature that facilitates the movements of organisms and
genetic flow (Taylor et al., 1993). It allows recovery from population decline, increase
in genetic diversity, long-term persistence and adaptation to climate change and habi-
tat fragmentation (McRae et al., 2012). Functional connectivity depends both on the
spatial configuration of the landscape and on the dispersal ability of the focal species.
For instance, the same forest landscape can be perceived as connected to a species
with high mobility, but as fragmented to a different species showing more restricted
movements (Baguette and Van Dyck, 2007).
Poor connectivity and isolation of populations have been identified as key factors to
the conservation of animal species with low dispersal ability (Thomas, 2000). There-
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
w y
fore, restoration of connectivity is a major concern in conservation practice. However,
identifying critical areas for the connectivity of a target population is not straightfor-
ward. It requires data on the distribution and demography of related (meta) popula-
tions, on the structural and spatial configuration of the landscape, and on the
mechanisms that drive individuals to move across that landscape (Hanski, 1998). In
many practical contexts, this kind of data is not readily available, so it is necessary to
use remote sensing (Corbane et al., 2015) and modelling techniques, that provide
spatially explicit patterns to design mitigation interventions and to address the lack of
connectivity (Jordán et al., 2003; Kajtoch et al., 2014). Currently, this is the only fea-
sible approach to many conservation problems, despite debates over the reliability
and applicability of modelling in the practical management of ecosystems.
Saproxylic beetles form a functional group of insects that depend on deadwood. Forest
patches with mature trees and decaying substrates harbour a high diversity of sapro-
xylics (Grove, 2002; Lachat et al., 2012; Seibold et al., 2015). Many species and the
whole community are good examples of occurrence constrained by habitat fragmen-
tation (Bouget et al., 2014; Jeppsson and Forslund, 2014; Janssen et al., 2016), lack
of high-resolution distribution knowledge (Buse et al., 2007) and an “active manage-
ment/reserve-designation” approach to conservation (D’Amen et al., 2013; Sebek et
al., 2013; Gran and Götmark, 2019; Karpi ski et al., 2021). Although various European
LIFE projects have targeted conservation of saproxylic beetle communities (European
Commission, 2012), initiatives explicitly aimed at increasing functional connectivity
for this group are rare in Europe, but have been recently demanded (Houston et al.,
2020). Single species systematic conservation planning has seldom been applied, as
opposed to multi-species or biodiversity planning (Mizsei et al., 2021).
Rosalia alpina (Linnaeus 1758) is a cerambycid saproxylic beetle, classified as “vulne-
rable” by the IUCN Red List (World Conservation Monitoring Centre, 1996) and as
“least concern” at European level (Horák et al., 2010). It is included in annexes II and
IV of Directive 92/43/CEE. Its habitat consists of Atlantic and continental broadleaved
woodlands (beech Fagus sylvatica forests in particular) with veteran trees (Cizek et al.,
2009; Drag et al., 2018). The long-term reduction and fragmentation of these wood-
lands, due mainly to transformation into agriculture fields, have driven the current
species geographic distribution in Europe and the Iberian Peninsula, with extensive
gaps and remnant populations in mountain ranges (Nieto and Alexander, 2010; Viño-
las and Vives, 2012; Bosso et al., 2013). Besides, intensive forestry techniques and
the removal of dead wood have caused detrimental effects on the suitability of the
existing forest habitats, leading to decline of populations (Nieto and Alexander, 2010).
Because of the current dominance of regenerating woodlands and commercial plan-
tations at the landscape level over the European range of R. alpina, the isolation of
habitat patches harbouring mature trees and dead wood, and the low proportion of
observed long-distance (>0,5 km) dispersal movements (Gatter, 1997; Drag et al.,
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
2011), connectivity is a relevant conservation concern for the species (Duelli and Wer-
melinger, 2005; Drag et al., 2011; Viñolas and Vives, 2012). Therefore, restoring habi-
tat cores and/or connecting areas is a key issue to support gene flow across the
landscape (Bosso et al., 2013), but practitioners face the challenge of selecting the
most adequate locations and forest patches to maximise the effectiveness of such
strategy.
The identification of forest patches with suitable microhabitats (i.e. individual trees
large enough to maintain tree-related microhabitats; Kozák et al., 2023) for R. alpina,
in the absence of high resolution and systematic field data on spatial distribution,
requires the development of robust probabilistic models (Jansson et al., 2009; Bosso
et al., 2017) based, if possible, on local knowledge of the species ecology (Russo et
al., 2010). An alternative strategy, based on the broad selection of corridors among
locations with known occurrence of the species, is not readily applicable in study areas
-like ours, see below-, where the available occurrence data has not been systematically
collected (i.e., there are probably pseudo-absences) at the spatial scale which is rele-
vant for practitioners involved in actual management (i.e., the surface of legally clas-
sified forests or montes de utilidad pública, for which compulsory action plans are
drafted and implemented). Available records for R. alpina in the study area are con-
centrated in natural parks, where opportunistic observations and some studies
focused on saproxylic beetles have been carried out (Castro and Fernández, 2018);
e. g., only 8 georeferenced records are retrieved from the Global Biodiversity Infor-
mation Facility in Gipuzkoa (checklist dataset https://doi.org/10.15468/39omei
accessed via GBIF.org on 2023-08-08). The knowledge on the distribution of the
species is thus biased to sites which have been sampled and shows low resolution.
Classified forests in Gipuzkoa are relatively small on average (χ
−= 93.9 ha, s=294.1,
n=311). Finally, corridors designed at coarse scales are rarely applied in real conser-
vation due to practical problems (Boitani et al., 2007; Keeley et al., 2018), and this
is the case in the study area, where ecological corridors targeting an ideal “forest
species” were proposed 20 years ago (Gurrutxaga et al., 2010) but to date no restora-
tions have taken place.
Here, we use a modelling approach to identify priority areas for the conservation of a
regional population of R. alpina. The study had the following specific objectives: (1)
to identify forest patches holding adequate microhabitats for R. alpina at a spatial
resolution that is useful for forest planners, managers and practitioners (1:10,000
scale); (2) to identify core patches (“nodes” in terms of connectivity) for R. alpina,
potentially receiving immigrants or emitting dispersing individuals; and (3) to identify
connecting patches (“links” or “stepping-stones”) that would support dispersal
between nodes in R. alpina populations. A preliminary version of this paper was
presented at the 6th Spanish Forestry Conference (Fernández-García et al., 2013).
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Material and methods
Study area
The study area was the province of Gipuzkoa (Basque Country, northern Spain). It
comprises 1,978 km2and is located in the Atlantic biogeographic region of Europe.
Climate is temperate and humid to hyperhumid (c. 2,000 mm of annual rainfall).
Although it is densely populated and urbanized in the coastal fringe and in low-alti-
tude valleys, it has also mountain ranges (up to c. 1,550 m asl) where woodland
coverage is dominant. Overall, 51.9 % of the province is covered by forest, and
about 51.6 % of this surface consists of coniferous commercial plantations (mostly
Monterrey pine Pinus radiata, larch Larix spp., Douglas fir Pseudotsuga menziessi and
Corsican pine P. nigra salzmannii). Beech forest surface reaches 15,315 ha, scattered
mainly in the southern and eastern districts of the province (Dirección General
de Desarrollo Rural y Política Forestal, 2013). R. alpina occurrence in Gipuzkoa is asso-
ciated to beech forests at mountain ranges (Castro and Fernández, 2018).
Habitat selection model
First, we applied two predictive models of R. alpina habitat selection at the tree level,
previously published for Gipuzkoa by Castro et al. (2012) and Castro and Fernández
(2016), using binary logistic regression. The model based on the occurrence of fresh
emergence holes (i.e. indicating reproduction) showed that the effective development
of the larvae preferably takes place in standing beech trees with dead parts, located
inside or at the edges of clearings, exposed to direct, regular sunlight and sheltered
from humid winds. The model based on the occurrence of adults suggested that they
preferred dead, standing or fallen beeches and large clearings with no preferred
orientation (see Castro and Fernández, 2016 for details on both models). Following
quantitative criteria and thresholds suggested by Castro et al. (2012), the ecological
niche of R. alpina in Gipuzkoa at the tree level was parameterised by applying five
sequential conditions: (1) beech, (2) dead standing tree, (3) tree diameter at 1.30 m
>50 cm, (4) dry exposure (other than N-NE-NW-W), and (5) tree inside of or at the
edge of a clearing >75 m2.
Identification of habitat patches
The following available cartographic tools were used: (1) a vegetation map of the
National Forest Inventory at scale 1:10,000 (Villanueva, 2009); (2) a map of exposures
derived from a digital elevation model, built using a raster dataset with a 25x25-m
pixel size (Gobierno Vasco, 2001) and a multiband Light Detection and Ranging
(LiDAR) image derived from a flight conducted in 2012, with a pixel size of 20x20-m
(Diputación Foral de Gipuzkoa, 2013). We generated a mask with the features “tree
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
species” (restricted to beech), “exposure” and “tree density”. To address the diffe-
rences in scale and overlay, we added a 30 m buffer to the resulting forest patches,
and then pixels with total or partial adequate exposure were selected.
Furthermore, a point GIS layer of dead or dying beeches with normal diameter (DBH)
>50 cm inventoried in the field was used to produce the model. This layer comprised
63 trees, occupying 56 pixels (20x20 m) distributed throughout the whole study area.
A supervised classification of the LiDAR image was performed by the spline interpo-
lation method, restricted to three standard deviations from the mean of the original
set of 56 20x20 m pixels. As a result, a total of 5,141 pixels were initially classified
(classification 1). Subsequently, a field survey in representative areas was conducted
to assess the accuracy of this classification. In total, 64 correct and 145 incorrect pixels
were assessed. The pixels correctly classified in the field visits were divided into 11
independent subgroups according to their location and proximity, and a new super-
vised classification of the LiDAR image was performed by the spline interpolation
method at two standard deviations from the means of each of the 11 subgroups
(classification 2). In addition, a supervised classification of the original LiDAR image
was obtained at two standard deviations from the mean of the 145 pixels that were
incorrectly classified (classification 3). Finally, the pixels in classification 2 that did not
coincide with classification 3 were accepted into the model. The result consisted of
3,062 pixels that indicated dead or decaying beeches, fitting the topographic,
physiognomic, and structural conditions described above.
Once the LiDAR image was classified, we assessed the reliability of the modelling
through the confusion matrix and the Kappa estimator. Field verification was con-
ducted for 5% of the pixels, which were randomly chosen, noting that 60.8% inclu-
ded dead or decaying beeches in adequate exposures, and were located inside or at
the edge of a clearing. This level of accuracy was considered sufficient to address the
study objectives.
Assessment and prioritisation of patches
We used Conefor Sensinode 2.5.8, a software that ranks habitat patches based on
their importance for connectivity. The software includes a graphing theoretical algo-
rithm that has been widely and successfully applied in Landscape Ecology research
(Urban and Keitt, 2001; Pascual-Hortal and Saura, 2006a, b). Graphs are mathematical
structures composed of nodes and links that contain descriptive attributes: area or
quality of habitat in the case of nodes and probability of dispersion between two
nodes through the link, estimated from Euclidean or functional distances. Conefor
Sensinode generates indices that integrate the intrinsic characteristics of the nodes
and the topological relationships among them.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
One of the indices is Equivalent Connected Area (ECA), which estimates the overall
degree of connectivity of a given landscape. It is defined as the surface of a hypothe-
tical continuous habitat patch (therefore fully connected, regardless of its form), that
would match the same connectivity probability assessed from the set of patches under
study (Saura and Rubio, 2010; Saura et al., 2011a). ECA is calculated using the
following expression:
where aiand ajare the areas of patches i and j. ECA has surface units, which facilitates
its interpretation. Its value will never be smaller than the area of the largest existing
patch in the set under study, and it will coincide with the total habitat area when
there is a single patch. Even when several differentiated patches exist, the probability
of moving through the available links is maximum for all tessellation edge pairs
(p*ij=1). The global connectivity of the population in the study area is expressed by
the relationship between ECA and the total area of available habitat, considering the
sum of the surface of the pixels identified by the modelling procedure.
The Probability of Connectivity (PC) index describes the probability of direct dispersion
between each pair of nodes and is sensitive to changes that may affect the connectivity
and availability of habitat (Saura and Pascual-Hortal, 2007). PC ranges from 0 to 1 and
is defined as the probability that two organisms, randomly situated in the landscape,
are found in nodes that are interconnected. Its mathematical expression is as follows:
where n is the number of nodes in the landscape; aiand ajare attributes of nodes i
and j (surface, quality, etc.); ALis the maximum value of the attribute of the landscape;
and p*ij is the maximum probability product between nodes i and j. The hierarchy of
landscape elements by their contribution to the availability and connectivity of the
entire habitat was obtained from the percentage of change of PC (dPCk), which is
generated by removing each k element of the landscape (Keitt et al., 1997; Urban
and Keitt, 2001; Saura and Pascual-Hortal, 2007):
The pixels accepted after the modelling were grouped into polygons if the euclidean
distance between them, calculated with ArcGIS 10.0, was <
–150 m (Fig. 1). We
assumed this as the dispersal distance for R. alpina, according to empirical data (Drag
et al. 2011; 80 % of recaptures of marked individuals occurred <
–150 m). The calcu-
lation of dPCkwas performed with Conefor Sensinode by applying a probabilistic
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
model of connectivity after a negative exponential function was fitted to the empirical
data of maximum likelihood dispersion described by Drag et al. (2011):
From a functional point of view, the importance of an element k (node, link) in the
landscape was divided into three fractions, after Saura and Pascual-Hortal (2007):
The fraction dPCintra is the amount of habitat provided by the node itself due to its
size (intrapatch connectivity). The fraction dPCflux is defined as the flow through the
connections that affect the node when it is the source or destination of that flow.
Finally, the fraction dPCconnector expresses the contribution of the node to the connec-
tivity among the network of nodes, functioning as a connecting or linking element.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Fig. 1.- Example of selected 20x20-m pixels in a forest massif in Gipuzkoa, indicating potential habitat
for Rosalia alpina (blue dots), and of patches aggregating pixels less than 150 m apart (polygons in
red). These pixels were classified applying a habitat selection model at tree resolution to a LiDAR image
(see text for details).
Fig. 1.- Ejemplo de píxeles de 20x20 m seleccionados en un macizo forestal de Gipuzkoa, que indican
habitat potencial para Rosalia alpina (puntos azules), y de polígonos que enlazan píxeles separados
por menos de 150 m (en rojo). Los píxeles se clasificaron aplicando un modelo de selección de habitat
a escala de árbol a una imagen LiDAR (ver texto para más detalles).
Using the quintiles from the scores obtained for the polygons, these were classified
as “core areas” or “connecting areas”. In the logical framework of the Landscape
Ecology theory, the former are “nodes” and the latter “stepping-stones”. The follo-
wing criteria were applied: (1) core areas were those that, in addition to having a high,
medium-high or medium value of the intra fraction of the PC index, also showed
medium and medium-high values of the flux fraction; (2) connecting areas were those
with high, medium-high and medium values of the connector fraction, or those with
medium-low values of the connector fraction and medium or medium-low values of
the intra fraction.
Results
We identified 380 polygons (forest patches) of different size, containing structural ele-
ments at the tree level required for R. alpina (Fig. 2). The largest number of patches
and the most important in terms of population connectivity, as ranked by the dPCk
model, were concentrated in south-eastern Gipuzkoa (Fig. 3). This area corresponds
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Fig. 2.- Geographical occurrence of 20x20-m pixels indicating potential habitat for Rosalia alpina in
Gipuzkoa (blue dots). The inset shows the location of the study area in south-western Europe.
Fig. 2.- Rango geográfico de los píxeles de 20x20 m que indican habitat potencial para Rosalia alpina
en Gipuzkoa (puntos azules). La ubicación del área de estudio en el suroeste de Europa se muestra en
el mapa pequeño.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Fig. 3.- Geographical distribution of polygons representing forest patches with structural elements for
Rosalia alpina in the study area (above; relevant special areas of conservation of the Natura 2000 net-
work are delineated), and example of their importance for the connectivity of the population in a forest
massif of Gipuzkoa, based on the dPCk index calculated with Conefor Sensinode software (below).
Fig. 3.- Distribución geográfica de los polígonos que representan parches de hábitat conteniendo ele-
mentos estructurales para Rosalia alpina en el área de estudio (arriba; se delimitan las zonas especiales
de conservación relevantes de la red Natura 2000), y ejemplo de su importancia para la conectividad
de la población en un macizo forestal de Gipuzkoa, según el índice dPCk calculado con la aplicación
Conefor Sensinode (debajo).
to the mountain ranges of Aizkorri-Aratz and Aralar, which are both classified as Spe-
cial Areas of Conservation (SAC) within the European Natura 2000 network. No other
polygons with such high values were identified in the rest of the study area.
The ratio between ECA and the total surface of the modelled habitat was 19.4, sug-
gesting that the global connectivity for the R. alpina population in the study area was
low. The largest contribution to the global connectivity of the population was the flux
fraction of PC (56.33%); many habitat patches were sources and receptors for dis-
persing individuals, therefore contributing to metapopulation dynamics at the lands-
cape scale. The intra fraction (37.23%) was less relevant, indicating the existence of
fewer patches with sufficient availability of microhabitats to maintain self-sustaining
subpopulations. The connector fraction (6.5%) was low, so few patches would be
able to act as links in the connectivity system. Only one of the patches in the con-
necting areas had a high value corresponding to the connector fraction. This suggests
deficiencies in the functionality as stepping-stones in most of the patches identified.
Discussion
Model usefulness and limitations
The fragmentation and isolation of the forest patches occupied by saproxylic beetles
play a major role in the long-term persistence of populations, precisely because of
their limited dispersal abilities (Ranius, 2006; Brunet and Isacsson, 2009; Janson et
al., 2009; Brin et al., 2016; Komonen and Müller, 2018), and this impact has been
also described for R. alpina (Drag et al., 2011). Addressing the connectivity issue by
modelling to guide conservation on-ground, especially when complete, systematic
data on species occurrence are not readily available, optimises the cost-efficient
balance of the action investments for flagship and threatened species (Dudley and
Vallauri, 2004; Mizsei et al., 2021), like R. alpina. The analysis and proposals in this
paper have few precedents regarding saproxylics conservation at the eco-regional level
(Nieto and Alexander, 2010).
The previous quantification of the species’ ecological niche in the study area (Castro
et al., 2012; Castro and Fernández, 2016), the information about relevant aspects of
the spatial ecology and demography (dispersal ability, stages and durations of life
cycles) of R. alpina (Drag et al., 2011) and the availability of remote-sensing data with
high resolution allowed us to identify patches of habitat that fit the combinations
with high probability of occurrence. This was the only feasible approach to achieve
the objective of this study and facilitated the classification of each patch from the
point of view of its connectivity function as node or stepping-stone.
The reliability of the ecological niche of R. alpina used here (from Castro et al., 2012;
Castro and Fernández, 2016) is high, because other studies on habitat preferences
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
published elsewhere in Europe had similar results (Russo et al., 2010; Campanaro et
al., 2017). The model by Russo et al. (2010) highlighted bark thickness as an influential
variable, but in the model for Gipuzkoa bark thickness was assumed to be closely cor-
related with the availability of forest gaps and insolation. Other independent qualita-
tive approaches to the habitat of R. alpina in Gipuzkoa (Pagola, 2011, unpublished
report), based on field observations in the Aiako Harria mountain range (NE of the
study area), are consistent with the importance of the discontinuities and the sun-
exposed clearings inside dense beech forests.
Unfortunately, the identification of deadwood from LiDAR and remote sensing data
is not possible yet, less said the characterisation of decay stage, which are fundamental
predictors of saproxylic occurrence. But other structural features, like volume, height
and DBH are surrogates for presence of tree-related microhabitats (Sebek et al., 2013;
Kozák et al., 2023). Given the forestry history of Gipuzkoa (Aragón, 2010), most exis-
ting beech trees that meet the DBH threshold are pollards. These trees have a response
profile in the LiDAR image characterised by maximum biomass at 2–3 feet off the
ground or a peak more pronounced near the apical zone. The field evaluation
endorsed the reliability of our spatially explicit model. LiDAR layers are increasingly
used for measuring forest structure across large areas, to encompass species
niches reliably (Müller and Brandl, 2009; Zellweger et al., 2014; Davison et al., 2023).
Moreover, the clustering of selected pixels to generate polygons analysable with
Conefor Sensinode improved the modelling performance, because the probability of
including favourable microhabitats within these polygons was increased.
Several limitations of the modelling approach have to be acknowledged. Our model
did not consider the connectivity with eventual populations at forest patches outside
of the study area. Secondly, the connectivity model has only considered the topological
(Euclidean distances) relationship among patches, without concerns about the actual
permeability of the matrix. However, at the scale of elements interconnected, this
matrix consists mainly of regenerating beech forest, so a substantial variability from
one relationship to the others is not expected.
R. alpina dispersion and performance of the model
Dispersion of R. alpina may be influenced by the species mobility (Komonen and
Müller, 2018) and this can be context-dependent (Russo et al., 2010; Drag et al., 2011;
Campanaro et al., 2017), but empirical data from the study area was not available.
Therefore, there may be concerns about the performance of the scale applied to pre-
dict habitat selection and presence of R. alpina in Gipuzkoa. However, it has been
suggested that low or slightly mobile saproxylics depending on ephemeral resources,
like deadwood, use the landscape reaching small patches of habitat in close proximity
within larger tracts of unsuitable forest matrix (Winiger et al., 2023).
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
The negative exponential function has been used to describe the distribution of the
probability of dispersal related to Euclidean distance in large-bodied saproxylic beetles
with low mobility (e. g. Osmoderma eremita; Ranius, 2006; Svensson et al., 2011).
Empirical dispersal data for R. alpina in central Europe were fitted to the function pro-
posed by Drag et al. (2011), with an estimation of a median dispersal distance of
161.5 m, close to the 150 m assumed by us to include most dispersal events. It is
highly likely that in the forests of Gipuzkoa average movement events are rather short
because Castro and Fernández (2016) found in our same study area that 200 m was
the maximum distance between the nearest occupied trees by R. alpina. Low-frequen-
cy, long-distance flights (up to 1.6 km; Drag et al., 2011; Rossi de Gasperis, 2015)
may play a role in ensuring gene flow in such spatial configurations. But to delineate
potential habitat patches, the average dispersal events are more adequate; therefore,
our 150 m threshold seems to adjust to this pattern of ecological use of the landscape.
Besides, in our study area, where old-growth habitat patches tend to be clum-
ped within beech forest matrix, dispersal through short-distance flights is probably
prevalent.
The identification of core and connecting areas for R. alpina in Gipuzkoa were con-
sistent with the expected pattern for a species showing low mobility across fragmen-
ted landscapes. Beech forests are concentrated in the south-eastern quadrant of the
study area (Dirección General de Desarrollo Rural y Política Forestal, 2013), so the iso-
lation by distance from the remaining beech patches throughout the rest of the ter-
ritory is extremely high. Moreover, within the larger beech tracts in the South-east,
favourable habitat patches with veteran trees are also fragmented, because of the
spatial and structural heterogeneity resulting from the history of forest exploitation
and the substitution of seminatural forests by plantations. Patches with connector
functionality were rarely identified. Overall, the modelling suggested a system in which
individuals within each cluster of near habitat patches have little potential difficulty in
moving between patches, but dispersal among forest tracts further apart, and the
establishment of metapopulation dynamics, proved unlikely.
Conservation implications
Reserve selection and retention techniques through sparing individual trees, groups
of trees or small forest patches from exploitation in managed forests have proved par-
ticularly beneficial for low o slightly mobile saproxylics (Bouget and Parmain, 2016;
Winiger et al., 2023). In this context, identification and ranking small forest patches
comprising clusters of habitat-trees according to their relevance as nodes and links,
can facilitate practitioners the restoration and enhancement of the connectivity for
R. alpina, because it provides recommendations as to where the interventions will
maximize the results.
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Although the performance of remote sensing data and complex modelling approaches
to aid conservation planning is open to discussion (e.g. Vatka et al., 2014), given the
lack of complete and systematic data on distribution, demography, and dispersion of
R. alpina in the study area, modelling was the logistically most feasible tool to imple-
ment scientifically sound conservation (Bosso et al., 2013). For continental or region-
al-scale management of biodiversity, models have been used successfully to detect
core nuclei or gaps in species ranges, recognise or delimitate reserves, corridors, or
stepping-stone interconnections (Bosso et al., 2017), as well as to propose sites where
restoration actions should be performed (Fitzgerald et al., 2008; Márcia et al., 2010).
Improving the connectivity of the whole population includes modifying some attribu-
tes of those patches whose functionality is compromised. This is the most efficient
strategy, because it is necessary to increase the ECA over the net increase in habitat
surface to achieve improved connectivity. In other words, the restoration of habitat
patches in sites lacking possibilities of being colonised would be useless; and restora-
tion in patches that are already connected and part of a metapopulation system would
be redundant.
The properties of the landscape must be considered for efficient habitat improvement
(Mortelliti, 2013). In our study area, patch size is the most manageable attribute,
because the topological position is obviously not an accessible factor, and we lack
precise information about the intrinsic structural quality of each patch. According to
Brunet and Isacsson (2009) and Sverdrup et al. (2010), treatments to improve the
availability of microhabitats are more effective if performed at the periphery of selected
patches, rather than at locations randomly distributed in the forest landscape, so that
the strategy of identifying specific areas for restoration is fully justified.
Management recommendations
The measures most frequently applied to improve the resilience of saproxylic beetle
populations include forestry interventions that actively bring stand structure and
woodland physiognomy closer to optimal conditions. These conditions are often set
on certain types of semi-natural landscapes, like open forests or even pasturelands
and parks with big trees in low densities (Jonsell, 2011; Widerberg et al., 2012;
Ramírez et al., 2014), along with increasing the volume of dead or decaying wood
(Cavalli and Mason, 2003; Davies et al., 2008). Such strategies are based on the the-
oretical enhancing effect of the increase in the availability of microhabitats provided
by veteran and damaged trees, in terms of foraging and breeding resources (Müller
et al., 2014).
Based on the relevant scientific literature, management guidelines can be drawn that
are applicable to the identified nodes and stepping-stones (Gossner et al., 2013b).
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Thus, (1) the optimal patches (nodes) should be strictly preserved to ensure long-term
maintenance of the current connectivity, and (2) interventions should be made on
those patches that could potentially act as stepping-stones, but whose current con-
ditions only allow poor functionality. For the former, the general approach consists of
adopting a protection regime, maintaining clearings, and limiting woodland densifi-
cation and shrub encroachment.
As for the stepping-stones, silvicultural management should retain in the patch all
dead or decaying wood that naturally or artificially occurs. A beech forest patch with
less than 20–30 m3of dead wood per ha or less than 4–8% of the total volume of
timber would not provide sufficient microhabitats for the saproxylic fauna (Dudley
and Vallauri, 2004). The average volume of dead wood in the beech forests of
Gipuzkoa is 9.0 m3/ha (Alberdi et al., 2012), far below the one measured in European
forest reserves (Christensen et al., 2005). In the periphery of the patch, active inter-
ventions should be planned to promote dead or rotting structures (branches, logs, or
snags) in the forest canopy and the understory (Cavalli and Mason, 2003; Sebek et
al., 2013). To increase sunlight exposure, clearings should be opened, or the density
of the patch should be reduced by cutting or uprooting a few trees, either leaving the
trunk on the ground or resting it over neighbouring trees (Cavalli and Mason, 2003;
Koch et al., 2012).
Concluding remarks
This study presents a feasible approach to identifying high important forest patches
for the conservation of R. alpina, taking advantage of remote sensing data combined
with modelling habitat selection at tree scale. Although these results are only appli-
cable to our own study area, we have shown that connectivity models can inform
conservation planning by making reliable predictions and therefore allowing targeted
actions. Reserve selection using expert judgement may lead to biased or ineffective
options (Saura et al., 2011b). Our emphasis on functional connectivity is justified,
because fragmentation of old-growth habitat patches in landscapes dominated by
managed, commercial forest stands is a major conservation concern for saproxylic
species. Testing in the field the actual performance of the models in terms of real
occurrence of R. alpina, and the improvements in functional connectivity and conser-
vation prospective for the species achieved with the proposed implementation of
regulations and management actions, would be a logical follow-up objective for the
future.
Acknowledgements
This work was supported by the project LIFE08/NAT/ES/00075, co-funded by the Euro-
pean Commission and the Provincial Council of Gipuzkoa. A. Cantero, A. Castro, L.
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Cizek, L. Drag, M. de Francisco, S. Hardersen, F. Mason, I. Mondragón, S. Pagola,
P. Riaño and S. Saura kindly provided expert advice. M. Méndez and E. Micó made
valuable comments on the submitted draft. P. Aramendi collaborated in the fieldwork.
Bibliography
Alberdi, I., Hernández, L., Saura, S., Barrera, M., Gil, P., Condés, S., Cantero, A., Sandoval,
V. J., Vallejo, R., Cañellas, I., 2012. Estimación de la biodiversidad forestal en el Tercer Inven-
tario Forestal Nacional. País Vasco. Ministerio de Agricultura, Alimentación y Medio
Ambiente, Madrid.
Aragón, A., 2010. Guided pollards in the Basque Country (Spain) during the early modern
ages. Landscape Archaeological Ecology 8, 14-22.
Baguette, M., Van Dyck, H., 2007. Landscape connectivity and animal behavior: functional
grain as a key determinant for dispersal. Lands. Ecol. 22, 1117-1129.
Boitani, L., Falcucci, A., Maiorano, L., Rondinini, C., 2007. Ecological networks as concep-
tual frameworks or operational tools in conservation. Conserv. Biol. 21, 1414-1422.
Bosso, L., Rebelo, H., Garonna, A. P., Russo, D., 2013. Modelling geographic distribution
and detecting conservation gaps in Italy for the threatened beetle Rosalia alpina. J. Nat.
Conserv. 21, 72-80.
Bosso, L., Smeraldo, S., Rapuzzi, P., Sama, G., Garonna, A. P., Russo, D., 2018. Nature pro-
tection areas of Europe are insufficient to preserve the threatened beetle Rosalia alpina
(Coleoptera: Cerambycidae): evidence from species distribution models and conservation
gap analysis. Ecol. Entomol. 43, 192-203.
Bouget, C., Larrieu, L., Brin, A., 2014. Key features for saproxylic beetle diversity derived
from rapid habitat assessment in temperate forests. Ecol. Indic. 36, 656-664.
Bouget, C., Parmain, G., 2016. Effects of landscape design of forest reserves on saproxylic
beetle diversity. Conserv. Biol. 30, 92-102.
Brin, A., Valladares, L., Ladet, S., Bouget, C., 2016. Effects of forest continuity on flying
saproxylic beetle assemblages in small woodlots embedded in agricultural landscapes. Bio-
div. Conserv. 25, 587-602.
Brunet, J., Isacsson, G., 2009. Restoration of beech forest for saproxylic beetles. Effects of
habitat fragmentation and substrate density on species diversity and distribution. Biodivers.
Conserv. 18, 2387-2404.
Buse, J., Schröder, B., Assmann, T., 2007. Modelling habitat and spatial distribution of an
endangered longhorn beetle. A case study for saproxylic insect conservation. Biol. Conserv.
137, 372-381.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Campanaro, A., Redolfi, L., Hardersen, S., Antonini, G., Chiari, S., Cini, A., Mancini, E.,
Mosconi, F., Rossi, S., Solano, E., Bologna, M. A., Peverieri, G. S., 2017. Guidelines for the
monitoring of Rosalia alpina. Nat. Conserv. 20, 165-203.
Castro, A., Fernández, J., 2016. Tree selection by the endangered beetle Rosalia alpina in
a lapsed pollard beech forest. J. Insect Conserv. 20, 201-214.
Castro, A., Fernández, J., 2018. Coleópteros saproxílicos de interés comunitario en espacios
Red Natura 2000 en Gipuzkoa (norte de España). Munibe, Cienc. nat. 66, 21-57.
Castro, A., Martínez de Murguía, L., Fernández, J., Casis, A., Molino, F., 2012. Size and
quality of wood used by Rosalia alpina (Linnaeus, 1758) (Coleoptera: Cerambycidae) in
beech woodlands of Gipuzkoa (northern Spain). Munibe, Cienc. nat. 60, 77-100.
Cavalli, R., Mason, F. (eds.), 2003. Techniques for re-establishment of dead wood for
saproxylic fauna conservation. Gianluigi Arcari, Mantova.
Cizek, L., Schlaghalamersky, J., Borucky, J., Hauck, D., Helesic, J., 2009. Range expansion
of an endangered beetle: Alpine Longhorn Rosalia alpina (Coleoptera: Cerambycidae) spre-
ads to the lowlands of Central Europe. Entomol. Fenn. 20, 200-206.
Christensen, M., Hahn, K., Mountford, E. P., Odor, P., Standovár, T., Rozenbergar, D., Diaci,
J., Wijdeven, S., Meyer, P., Winter, S., Vrska, T., 2005. Dead wood in European beech (Fagus
sylvatica) forest reserves. For. Ecol. Manag. 210, 267-282.
Corbane, C., Lang, S., Pipkins, K., Alleaume, S., Deshayes, M., García, V. E., Strasser, T.,
Vanden Borre, J., Toon, S., Michael, F., 2015. Remote sensing for mapping natural habitats
and their conservation status. New opportunities and challenges. Int. J. Appl. Earth Observ.
Geoinf. 37, 7-16.
D’Amen, M., Bombi, P., Campanaro, A., Zapponi, L., Bologna, M. A., Mason, F., 2013. Pro-
tected areas and insect conservation: questioning the effectiveness of Natura 2000 network
for saproxylic beetles in Italy. Animal Conserv. 16, 370-378.
Davies, Z., Tyler, C., Stewart, G., Pullin, A. S., 2008. Are current management recommen-
dations for saproxylic invertebrates effective? A systematic review. Biodivers. Conserv. 17,
209-234.
Davison, C. W., Assmann, J. J., Normand, S., Rahbek, C., Morueta-Holme, N., 2023. Vege-
tation structure from LiDAR explains the local richness of birds across Denmark. J. Animal
Ecol. 92, 1332-1344.
Diputación Foral de Gipuzkoa, 2012. LiDAR data for the province of Gipuzkoa. Available
from: http://b5m.gipuzkoa.net/web5000/es/cartoteca/lidar-gipuzkoa/ [accessed 1 July
2020].
Dirección General de Desarrollo Rural y Política Forestal, 2013. Cuarto Inventario Forestal
Nacional. Ministerio de Agricultura, Alimentación y Medio Ambiente, Madrid.
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Drag, L., Hauck, D., Pokluda, P., Zimmermann, K., Cizek, L., 2011. Demography and dis-
persal ability of a threatened saproxylic beetle: a mark-recapture study of the Rosalia Longi-
corn (Rosalia alpina). PLOS One 6, e21345.
Drag, L., Hauck, D., Rican, O., Schmitt, T., Shovkoon, D. F., Godunko, R. J., Curletti, G.,
Cizek, L., 2018. Phylogeography of the endangered saproxylic beetle Rosalia longicorn,
Rosalia alpina (Coleoptera, Cerambycidae), corresponds with its main host, the European
beech (Fagus sylvatica, Fagaceae). J. Biogeogr. 45, 2631-2644.
Duelli, P., Wermelinger, B., 2005. Der Alpenbock (Rosalia alpina). Ein seltener Bockkäfer
als Flaggschiff Art. Merkblatt für die Praxis 39, 1-8.
Dudley, N., Vallauri, D., 2004. Deadwood-living forests. The importance of veteran trees
and deadwood to biodiversity. World Wildlife Fund, Gland.
European Commission, 2012. LIFE and invertebrate conservation. Publication Office of the
European Union, Luxembourg.
Fernández-García, J. M., Sesma, J., Moreno, E., Cantero, A., Riaño, P., De Francisco, M.,
Mondragón, I., 2013. Modelización e incremento de la conectividad para el coleóptero
saproxílico Rosalia alpina en Gipuzkoa. In: Actas del 6º Congreso Forestal Español, Sociedad
Española de Ciencias Forestales, Pontevedra.
Fitzgerald, J. A., Thogmartin, W. E., Dettmers, R., Jones, T., Rustay, C., Ruth, J. M., Thomp-
son, F. R., Will, T., 2008. Application of models to conservation planning for terrestrial birds
in North America. In: Millspaugh, J. M., Thompson, F. R. (Eds.), Models for planning wildlife
conservation in large landscapes, 593-624. Academic Press, Burlington.
Gatter, W., 1997. Förderungsmöglichkeiten für den Alpenbock. Allgemeine Forst Zeitschrift
für Waldwirtschaft und Umweltvorsorge 213, 1305-1306.
Gobierno Vasco, 2001. Digital elevation model for the Basque Autonomous Community.
Available from: www.geoeuskadi.eus [accessed 1 July 2020].
Gossner, M. M., Floren, A., Weisser, W. W., Linsenmair, K. E., 2013a. Effect of dead wood
enrichment in the canopy and on the forest floor on beetle guild composition. For. Ecol.
Manag. 302, 404-413.
Gossner, M. M., Lachat, T., Brunet, J., Isacsson, G., Bouget, C., Brustel, H., Brandl, R., Weiss-
er, W. W., Müller, J., 2013b. Current near-to-nature forest management effects on func-
tional trait composition of saproxylic beetles in beech forests. Conserv. Biol. 27, 605-614.
Gran, O., Götmark, F., 2019. Long term experimental management in Swedish mixed
oak rich forests has a positive effect on saproxylic beetles after 10 years. Biodivers. Conserv.
28, 1451-1472.
Grove, S. J., 2002. Saproxylic insect ecology and the sustainable management of forests.
Annual Review Ecol. Syst. 33, 1-23.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Gurrutxaga, M., Lozano, P. J., Del Barrio, G., 2010. GIS-based approach for incorporating
the connectivity of ecological networks into regional planning. J. Nat. Conserv. 18, 318-
326.
Hanski, I., 1998. Metapopulation dynamics. Nature 396, 41-49.
Horák, J., Tezcan, S., Micó, E., Schmidl, J., Petrakis, P., 2010. Rosalia alpina (Europe assess-
ment). The IUCN Red List of Threatened Species 2010: e.T19743A9009045.
Houston, J., Aramburu, M. J., Velghe, D., 2020. LIFE and Invertebrates. Stepping up to the
challenges: conclusions of the 2018 LIFE Platform Meeting on invertebrates and 2019 ex-
post visits to closed projects. Unpublished report for the European Commission.
Jansson, N., Bergman, K. O., Jonsell, M., Milberg, P., 2009. An indicator system for iden-
tification of sites of high conservation value for saproxylic oak (Quercus spp.) beetles in
southern Sweden. J. Insect Conserv. 13, 399-412.
Jeppsson, T., Forslund, P., 2014. Species’ traits explain differences in red list status and
long term population trends in longhorn beetles. Anim. Conserv. 17, 332-341.
Jordán, F., Báldi, A., Orci, K. M., Rácz, I., Varga, Z., 2003. Characterizing the importance
of habitat patches and corridors in maintaining the landscape connectivity of a Pholidoptera
transylvanica (Orthoptera) metapopulation. Landsc. Ecol. 18, 83-92.
Jonsell, M., 2011. Old park trees as habitat for saproxylic beetle species. Biodivers. Conserv.
21, 619-642.
Kajtoch, L., Mazur, M., Kubisz, D., Mazur, M. A., Babik, W., 2014. Low effective population
sizes and limited connectivity in xerothermic beetles: implications for the conservation of
an endangered habitat. Anim. Conserv. 17, 454-466.
Karpi ski, L., Maák, I., Wegierek, P., 2021. The role of nature reserves in preserving saproxy-
lic biodiversity: using longhorn beetles (Coleoptera: Cerambycidae) as bioindicators. Eur.
Zool. J. 88, 487-504.
Keeley, A., Basson, G., Cameron, D. R., Heller, N. E., Huber, P. R., Schloss,C. A., Thorne, J.
H., Merenlender, A. M., 2018. Making habitat connectivity a reality. Conserv. Biol. 32,
1221-1232.
Keitt, T. H., Urban, D. L., Milne, B. T., 1997. Detecting critical scales in fragmentised
landscape. Conserv. Ecol. 1, 4-6.
Koch, M., Ranius, T., Drobyshev, I., Nilsson, U., Lindbladh, M., 2012. Increased openness
around retained oaks increases species richness of saproxylic beetles. Biodivers. Conserv.
21, 3035-3059.
Komonen, A., Müller, J., 2018. Dispersal ecology of dead wood organisms and connectivity
conservation. Conserv. Biol. 32, 535-545.
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Kozák, D., Svitok, M., Zemlerová, V., Mikoláš, M., Lachat, T., Larrieu, L., Paillet, Y., Buech-
ling, A., Ba e, R., Keeton, W. S., Vítková, L., Begovi , K., ada, V., Dušátko, M., Feren ík,
M., Frankovi , M., Gloor, R., Hofmeister, J., Janda, P., Kameniar, O., Kní , T., Majdanová, L.,
Mejst ík, M., Pavlin, J., Ralhan, D., Rodrigo, R., Roibu, C.-C., Synek, M., Vostarek, O., Svo-
boda, M. 2023. Importance of conserving large and old trees to continuity of tree-related
microhabitats. Conserv. Biol. 37, e14066.
Lachat, T., Wermelingera, B., Gossner, M. M., Bussler, H., Isacsson, G., Müller, J., 2012.
Saproxylic beetles as indicator species for dead-wood amount and temperature in European
beech forests. Ecol. Indic. 23, 323-331.
Márcia, A., Real, R., Vargas, J. M., 2010. Use of coarse-resolution models of species’ dis-
tributions to guide local conservation inferences. Conserv. Biol. 24, 1378-1387.
McRae, B. H., Hall, S. A., Beier, P., Theobald, D. M., 2012. Where to restore ecological
connectivity? Detecting barriers and quantifying restoration benefits. PLOS One 7,
e52604.
Mizsei, E., Zabolcs, S., Szabó, L., Boros, Z., Mersini, K., Roussos, S. A., Dimaki, M., Ioannidis,
Y., Végvári, Z., Lengyel, S., 2021. Determining priority areas for an endangered cold-adap-
ted snake on warming mountaintops. Oryx 55, 334-343.
Mortelliti, A., 2013. Targeting habitat management in fragmented landscapes: a case study
with forest vertebrates. Biodivers. Conserv. 22, 187-207.
Müller, J., Brandl, R., 2009. Assessing biodiversity by remote sensing in mountainous ter-
rain: the potential of LiDAR to predict forest beetle assemblages. J. Appl. Ecol. 46, 897-
905.
Müller, J., Jarzabek, A., Bussler, H., Gossner, M. M., 2014. Hollow beech trees identified
as keystone structures for saproxylic beetles by analyses of functional and phylogenetic
diversity. Anim. Conserv. 17, 154-162.
Nieto, A., Alexander, K. A., 2010. European red list of saproxylic beetles. Publications Office
of the European Union, Luxembourg.
Pascual-Hortal, L., Saura, S., 2006a. Comparison and development of new graph-based
landscape connectivity indices: towards the prioritization of habitat patches and corridors
for conservation. Lands. Ecol. 21, 959-967.
Pascual-Hortal, L., Saura, S., 2006b. Integrating landscape connectivity in broad-scale forest
planning through a new graph-based habitat availability methodology: application to
capercaillie (Tetrao urogallus) in Catalonia (NE Spain). Eur. J. For. Res., 127, 23-31.
Ramírez, A., Micó, E., Marcos, M. A., Brustel, H., Galante, E., 2014. The “dehesa”, a key
ecosystem in maintaining the diversity of Mediterranean saproxylic insects (Coleoptera and
Diptera: Syrphidae). Biodivers. Conserv. 23, 2069-2086.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Ranius, T., 2006. Measuring the dispersal of saproxylic insects: a key characteristic for their
conservation. Popul. Ecol. 48, 177-188.
Rossi de Gasperis, S., 2015. Distribution patterns and population analysis of threatened
long-horn beetles in forest habitats of Central Italy. PhD Thesis. Università degli Studi Roma
Tre, Rome.
Russo, D., Cistrone, L., Garonna, A. P., 2010. Habitat selection by the highly endangered
long-horned beetle Rosalia alpina in Southern Europe: a multiple spatial scale assessment.
J. Insect Conserv. 15, 685-693.
Saura, S., Pascual-Hortal, L., 2007. A new habitat availability index to integrate connectivity
in landscape conservation planning: comparison with existing indices and application to a
case study. Lands. Urban Plan. 83, 91-103.
Saura, S., Rubio, L., 2010. A common currency for the different ways in which patches
and links can contribute to habitat availability and connectivity in the landscape. Ecography
33, 523-537.
Saura, S., Estreguil, C., Mouton, C., Rodríguez-Freire, M., 2011a. Network analysis to assess
landscape connectivity trends: application to European forests (1990-2000). Ecol. Indic.
11, 407-416.
Saura, S., Vogt, P., Velázquez, J., Hernando, A., Tejera, R., 2011b. Key structural forest con-
nectors can be identified by combining landscape spatial pattern and network analyses.
For. Ecol. Manag. 262, 150-160.
Sebek, P., Altman, J., Platek, M., Cizek, L., 2013. Is active management the key to the con-
servation of saproxylic biodiversity? Pollarding promotes the formation of tree hollows.
PLOS One 8, e60456.
Seibold, S., Brandl, R., Buse, J., Hothorn, T., Schmidl, J., Thorn, S., Müller, J., 2015. Asso-
ciation of extinction risk of saproxylic beetles with ecological degradation of forests in
Europe. Conserv. Biol. 29, 382-390.
Svensson, G. P., Sahlin, U., Björn, B., Larsson, M. C., 2011. Should I stay or should I go?
Modelling dispersal strategies in saproxylic insects based on pheromone capture and radio
telemetry: a case study on the threatened hermit beetle Osmoderma eremita. Biodivers.
Conserv. 20, 2883-2902.
Sverdrup, A., Skarpaas, O., Odegaard, F., 2010. Hollow oaks and beetle conservation: the
significance of the surroundings. Biodiver. Conserv. 19, 837-852.
Taylor, P. D., Fahrig, L., Henein, K., Merriam, G., 1993. Connectivity is a vital element of
landscape structure. Oikos 68, 571-573.
Thomas, C. D., 2000. Dispersal and extinction in fragmented landscapes. Proceedings of
The Royal Society London B 267, 139-145.
José María Fernández-García et al.
Munibe, Cienc. nat. 71, 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Urban, D., Keitt, T. H., 2001. Landscape connectivity: a graph-theoretic perspective. Ecology
82, 1205-1218.
Vatka, E., Kangas, K., Orell, M., Lampila, S., Nikula, A., Nivala, V., 2014. Nest site selection
of a primary hole-nesting passerine reveals means to developing sustainable forestry. J.
Avian Biol. 45, 187-196.
Villanueva, J. A. (Ed.), 2009. Tercer inventario forestal nacional. 1997-2007.
Guipúzcoa/Gipuzkoa. Ministerio de Medio Ambiente, Madrid.
Viñolas, A., Vives, E., 2012. Rosalia alpina. In: Galante, E., Micó, E. (Eds.), Bases ecológicas
preliminares para la conservación de las especies de interés comunitario en España: Inver-
tebrados. Ministerio de Agricultura, Alimentación y Medio Ambiente, Madrid.
Widerberg, M. K., Ranius, T., Drobyshev, I., Nilsson, U., Lindbladh, M., 2012. Increased
openness around retained oaks increases species richness of saproxylic beetles. Biodivers.
Conserv. 21, 3035-3059.
Winiger, N., Hendel, A.-L., Ganz, S., Zielwska-Büttner, K., Segelbacher, G., Braunisch, V.,
2023. Saproxylic beetles respond to habitat variables at different spatial scales depending
on variable type and species’ mobility: the need for multi-scale forest structure manage-
ment. Biodivers. Conserv. 32, 3355-3377.
World Conservation Monitoring Centre. 1996. Rosalia alpina. The IUCN Red List of
Threatened Species 1996: e.T19743A9009447.
Zellweger, F., Morsdorf, F., Purves, R. S., Braunishch, V., Bollmann, K., 2014. Improved
methods for measuring forest landscape structure: LiDAR complements field-based habitat
assessment. Biodivers. Conserv. 23, 289-307.
Rosalia alpina connectivity
Munibe, Cienc. nat. 71 2023 •Donostia/San Sebastián •ISSN 0214-7688 •eISSN 2172-4547
Fecha de recepción/ Date of reception: 23/09/2022
Fecha de aceptación / Date of acceptance: 03/09/2023
Editor Asociado / Associate editor: Alberto Castro
