BookPDF Available

Italian Ecological Network: the Role of Protected Areas in the Conservation of Vertebrates

Authors:

Figures

Content may be subject to copyright.
Luigi Boitani
Alessandra Falcucci
Luigi Maiorano
Alessandro Montemaggiori
University of Rome “La Sapienza”
Animal and Human Biology Department
Ministry of Environment
Nature Conservation Directorate
ITALIAN
ECOLOGICAL
NETWORK
THE ROLE OF THE
PROTECTED AREAS IN
THE CONSERVATION
OF VERTEBRATES
Luigi Boitani
Animal and Human Biology Department, University of Rome “La Sapienza”
Alessandra Falcucci
College of Natural Resources, Department of Fish and Wildlife Resources University of Idaho, Moscow (USA)
Animal and Human Biology Department, University of Rome “La Sapienza”
Luigi Maiorano
College of Natural Resources, Department of Fish and Wildlife Resources University of Idaho, Moscow (USA)
Animal and Human Biology Department, University of Rome “La Sapienza”
Alessandro Montemaggiori
Animal and Human Biology Department, University of Rome “La Sapienza”
Institute of Applied Ecology, Rome
ITALIAN
ECOLOGICAL
NETWORK
THE ROLE OF THE
PROTECTED AREAS IN
THE CONSERVATION
OF VERTEBRATES
September 2003
Recommended citation:
Boitani L., A. Falcucci, L. Maiorano & A. Montemaggiori. 2003.
National Ecological Network: the role of the protected areas in the conservation of vertebrates.
Animal and Human Biology Department, University of Rome “La Sapienza”,
Nature Conservation Directorate of the Italian Ministry of Environment, Institute of Applied Ecology.
All the data utilized in the study relating to the National Ecological Network for Vertebrates are currently available on-line at the following URL:
http://www.gisbau.uniroma1.it/REN/index.htm.
All photographs by Alessandro Montemaggiori ©.
Graphic design: Raffaella Gemma
Printed by: Tipografia Eurosia, Roma
Contents
1 Introduction 5
1.1 Ecological networks and biodiversity 6
2 The Protected Areas in Italy 8
2.1 Type and size of the Protected Areas in Italy 12
2.2 Physical and geographic characterization 14
3 Nature 2000 Network in Italy 18
4 The Ecological Network: a paradigm of conceptual reference 22
4.1 Methodological approaches 23
4.2 Animal Database 24
4.3 Environmental suitability models for the various species 25
4.4 Validation of suitability models 28
4.5 Ecological networks 29
4.6 Comparison between networks 30
5 Congruency analysis of Protected Areas based on number of vertebrate species 31
5.1 All species of vertebrates 32
5.2 Mammals 37
5.3 Birds 42
5.4 Reptiles 47
5.5 Amphibians 52
5.6 Fish 57
5.7 Endangered species 62
6 Congruency analysis of Protected Areas based on irreplaceability 67
6.1 All species of vertebrates 70
6.2 Mammals 75
6.3 Birds 78
6.4 Reptiles and Amphibians 80
6.5 Fish 82
7 Conclusions 84
8 Bibliography 87
ITALIAN
ECOLOGICAL
NETWORK
THE ROLE OF THE
PROTECTED AREAS IN
THE CONSERVATION
OF VERTEBRATES
Italian Regions and main mountain chains.
1 Introduction
In 1999, the Ministry of Environment (Nature Conser-
vation Directorate) approved a document outlining the
design parameters for the National Ecological Net-
work (REN – Rete Ecologica Nazionale) and defining
the structure and principal objectives of the system na-
tionwide. Basically, REN is an integrated programme
that aims to rebalance socio-economic development
trends within a framework of sustainable growth and
optimal biodiversity conservation and, as such, it takes
the form of a complex network of programmes relating
to widely varying sectors of the economy, culture, terri-
torial management and, of course, ecology and biodi-
versity management, particularly with regard to
species and types of habitat.
The new course of conservation policy follows the
main European Directives on nature conservation, de-
veloping them into a more holistic concept of the land
and its natural and human components. It is also inte-
grated with the renewed efforts of the Council of Eu-
rope to promote a more comprehensive, less fragment-
ed approach to territorial administration, recently lead-
ing to the adoption of the European Landscape Con-
vention. In short, this tendency aims to do more than
just emphasize the conservation of individual species
or protected areas: the focus has shifted onto a system-
atic policy involving all environmental components, in
close connection with European strategies.
The said document and related ministerial directives
must now find support among the various disciplines
for the definition of methods, guidelines and planning
characteristics, so as to create the basic structure of a
national network. In particular, it seems a good time to
set up a mechanism for calibrating the national eco-
logical network, with the specific aim, among other
things, of conserving Italian biodiversity.
Altogether, the protected areas in Italy (with the excep-
tion of the Sites of European Community Importance-
SCI - and Special Protection Areas- SPA -, which are
still at the approval stage) cover almost 11% of the
country; in view of their natural features, they can be
considered one of the most important components of a
potential ecological network dedicated to biodiversity
conservation (Figs. 1-3). Nevertheless, because of
their size and the criteria by which they were chosen,
Italy’s protected areas are not sufficient to satisfy the
requirements of biodiversity conservation. They can be
important nodes in an ecological network, as long as
they are considered within the context of the territory
as a whole and their role is verified within an environ-
mental infrastructure (protected areas and corridors)
that is also calibrated to take account of the biological
and ecological requirements of the various species
and habitats.
The present document sets out, in a concise and read-
able form, the results of an analysis of the content of
protected areas in Italy, particularly in terms of verte-
brate species. The aim of the study is to verify whether
the system of protected areas corresponds to the pat-
tern of biodiversity for vertebrates and to determine
what sort of action should be taken in order to render
the system more efficient in conserving this important
component of biodiversity. The study was carried out,
in response to a request from the Ministry of Environ-
ment, Nature Conservation Directorate, by the Animal
and Human Biology Department University of Rome
“La Sapienza”, and is based on data resulting from
the study completed in February 2002 on the defini-
tion of the National Ecological Network for the conser-
vation of vertebrates.
Introduction
1
5
Pollino National Park (Calabria – Southern Italy)
1.1 Ecological networks and biodiversity
Ecological networks are a conceptual tool of the ut-
most importance for nature conservation and sus-
tainable land use. Their theoretical foundation lies
deep in conservation biology and is based on the
obvious premises that all species, both animal and
vegetable, are distributed over the land with an ir-
regular pattern and that this discontinuity is due
chiefly to intrinsic natural factors, which may also be
aggravated by historical and human factors. The
distribution range of each species consists of a num-
ber of different areas, in which the density of the
species is not always the same. Under optimum con-
ditions, these areas are joined to one another by
corridors, forming a network. The links can be of
widely varying kinds, depending on the species in
question. They may consist of individuals that dis-
perse and move across the land following routes de-
termined to some extent by the suitability of the ter-
rain traversed, or they may be almost entirely inde-
pendent of the terrain because mobility is achieved
by aerial means (seeds, spores, birds, insects etc.).
It is therefore clear that the concept of ecological
network can find practical expression in completely
different ways, depending on the species taken into
consideration. The global ecological network, repre-
sented by overlaying the innumerable networks of
all animal and vegetable species, produces minute
fragmentation of the territory into tiny homogeneous
areas, representing the true - and theoretical - eco-
logical network existing in the country.
In practice, this “web of networks” can only be trans-
formed into an effective tool for managing the territory
by aggregating several similar areas so as to achieve
a level of detail that can be handled by the tools nor-
mally used for land use projects. To this end, it is help-
ful to reach the level of landscape characteristics, iden-
tifying the most homogeneous landscape units. While
this operation undoubtedly offers practical advan-
tages, however, it must not be seen as satisfying the re-
quirements of all species: there is no guarantee that a
network identified at a macroscopic level in this way
will help to conserve a significant proportion of animal
and vegetable species, nor is there any guarantee that
it will help to conserve endangered types of habitat.
An ecological network drawn up only on the basis of
landscape characteristics may therefore be entirely ir-
relevant with regard to the functional objectives set.
An alternative method, aimed at ensuring that an eco-
logical network represents a useful compromise be-
tween the requirements of the species and those of
land use, might be a network calibrated according to
the requirements of the species considered to be most
important for the conservation of numbers and for the
effectiveness of the systems; both these options must be
assessed according to the intended interpretation of
the global network. Once the network has been de-
fined, the homogeneous landscape units to be used for
land planning and management can also be defined.
Since it is not possible to consider the requirements of
all the species that inhabit a given area, the scope
Ecological networks and biodiversity
6
1.1
Little Bittern (Ixobrychus minutus)
Italian Tree Frog (Hyla intermedia)
1.1 Ecological networks and biodiversity
must necessarily be restricted to the species viewed as
critical, either because they are endangered or be-
cause they play a functional role within the ecological
systems. From a practical point of view, and in the
light of the various problems to be tackled by means
of the resulting network, the species can be divided in-
to various categories: a) keystone species, so called
on account of their important roles in the ecological
communities, b) umbrella species, so called because
they are generally to be found at the upper hierarchic
levels of the trophic chains and their conservation nec-
essarily brings with it that of the species found at low-
er levels, c) flag species, so called on account of their
ability to draw public attention and to facilitate conser-
vation activity.
In terms of species, biodiversity in Italy consists of over
57,000 animal species, while there are far fewer veg-
etable species. It would therefore seem right to give
priority to animal species, since their conservation
necessarily implies the conservation of the related veg-
etable systems on account of their position in the troph-
ic chains. And from a zoological point of view, verte-
brates undoubtedly occupy a “flag” role, to the point
of being only too often the sole reference for conserva-
tion policy.
7
Red Deers (Cervus elaphus)
The system of existing Protected Areas (PA) in Italy,
1,004 of them altogether (Gambino & Negrini,
2001), include all those set up by formal directives
issued by the State or the Regions. Not all the PA that
exist in practice are included in the Official List of
Protected Natural Areas (EUAP) which, according to
the 4th version issued by the Ministry of the
Environment and Territorial Protection, comprises
751 areas altogether.
Seventeen areas listed in EUAP (16 Protected Marine
Natural Areas and Marine Reserves and 1 Other
National Protected Natural Area), together with the
marine part of three National Parks (those of the
Maddalena Archipelago, the island
of Asinara and the Tuscan
Archipelago) were omitted from the
analyses carried out, because the
data used to construct the habitat
suitability models for Italian verte-
brates do not cover the Protected
Areas in question (for further details,
see par. 4.3.).
Altogether, 775 PA were taken into
consideration during the analysis
(Figs. 1-8): as compared with the
official list issued by the Ministry of
Environment, only one National
Reserve was omitted, besides the 17
areas mentioned above, because no relevant maps
could be found, but the study included 13 Other
Protected Natural Areas, 1 Regional Natural
Reserve and 28 Regional Natural Parks that do not
figure in the EUAP, either because the respective
regions did not apply to the Ministry
for registration or because these
areas were considered incongruent
with the conditions laid down by
Italian legislation.
All references to PA made in the
remainder of this document should be
taken as referring to the above-men-
tioned 775 areas, for which a geo-
referenced information system has
been set up comprising the most
recent data available as of
September 2002, specifying identifi-
cation details, institutional references
and territorial boundaries. The SCI
and SPA have been considered sepa-
rately (see par. 2.3), but portions of
them that fall within the protected
areas are included in this study.
2 The Protected Areas in Italy
The Protected Areas in Italy
2
8
Fig. 1. Number of Protected Areas per type in Italy.
Fig. 2. Number of Protected Areas of the various types per region.
2 The Protected Areas in Italy
9
Fig. 3. Protected Areas in Italy (775 areas).
Protected Areas
2 The Protected Areas in Italy
10
Fig. 4. National Parks in Italy.
Fig. 6. National Reserves in Italy.
Fig. 5. Regional Parks in Italy.
Fig. 7. Regional Reserves in Italy.
National Parks
Regional Parks
National Reserves
Regional Reserves
2 The Protected Areas in Italy
11
Fig. 8. Other Protected Natural Areas in Italy.
Fig. 10. SPA (Special Protection Areas) in Italy.
Fig. 9. SCI (Sites of European Community Importance) in Italy.
SPA
SCI
Other Protected Natural Areas
Altogether, the PA in Italy cover over 3 million
hectares
1
, corresponding to 11% of the national ter-
ritory (Fig. 11). They are distributed extremely irreg-
ularly among the regions (Fig. 12): in some regions,
25% or more of the territory is protected (e.g.
Abruzzi, Lombardy, Autonomous Province of
Bolzano, Campania), whereas in others less than 5%
of the territory is protected, as in the extreme cases
of Molise, 1.5%, and Sardinia, less than 1%.
If we analyse the overall distribution graph showing
the size of the Italian PA (Fig. 13), various interesting
aspects can be noted. For example, the PA have a
mean size of 4,352.5 ha, but the median size is only
265.4 ha, showing that 50% of the PA are smaller
than the median. The discrepancy between mean
and median area is due essentially to the consider-
able size of five National Parks, which cover a mean
area of over 100,000 ha and have a maximum size
(Pollino National Park) of over 183,000 ha
2
.
The arithmetical average of the size of the PA is
therefore heavily influenced by these five areas and
consequently does not give a correct picture of mean
conditions.
In fact, most of the PA cover a small area, at least
from the biological/ecological
point of view: 70% of them
cover less than 1,000 ha, 60%
less than 500 ha and 33% less
than 100 ha.
The smallest PA is the Sasso di
Preguda Regional Natural
Monument (Lombardy Region),
which covers 0.05 ha (accord-
ing to the EUAP drawn up by
the Ministry of Environment, it
has an area of 0 ha.) It is also
important to note that 9% of the
PA in Italy are smaller than 10
ha.
There is a notable degree of
overlapping between the vari-
ous kinds of PA (Table I). For
2.1 Type and size of the Protected Areas in Italy
Type and size of the Protected Areas in Italy
12
2.1
Fig. 12. Percentage of regional territory covered by the various types of Protected Areas. The percentages are only indicative,
because of overlapping between the various types of Protected Areas.
Fig. 11. Percentage of national territory covered by the various
types of Protected Areas. The percentages are only indicative,
because of overlapping between the various types of Protected
Areas.
2.1 Type and size of the Protected Areas in Italy
example, 47% of the area of National Reserves
(over 59,000 ha) falls within National Parks, while
4% (about 5,650 ha) falls within Regional Parks. As
far as the remaining types are concerned, the
degree of overlap is generally very low (on average
about 2%), but it should be emphasised that there
are a considerable number of cases in which a pro-
tected area falls entirely within another PA of a dif-
ferent kind.
13
1
The areas mentioned in the present publication were calculated on the basis of the digital maps used in the ArcGis environment: this means
that the areas given by the authors often do not coincide precisely with the official areas declared in the EUAP drawn up by the Ministry of
Environment. In certain cases, there is a considerable discrepancy between the official area of a PA and the area calculated using ArcGis: in
these cases, too, we have chosen to adopt the value calculated by using the maps, since this meant that it corresponded more closely to the
map situation.
2
It should be noted that, according to the EUAP drawn up the Ministry of Environment, the largest park is the Cilento and Vallo di Diano Na-
tional Park, which covers over 178,000 ha.
Fig. 13. Number of Protected Areas present in Italy subdivided
by classes of area (ha).
National Parks
Regional Parks
National Reserves
Regional Reserves
Other Protected Areas
SCI
SPA
0
0
0
0
0
0
0
0
059.980
188
640.756 638.230
0 5.647 3.795
597
866.812
535.802
59.980
5.647
555
1.155
79.901
86.616
188
3.795
555 1.009
152.071
49.206
0
597 1.155 1.009 41.995 14.146
640.756
866.812
79.901
152.071 41.995 1.391.213
638.230 535.802 86.616 49.206 14.146
0
1.391.213
National Parks
Regional Parks
National Reserves
Regional Reserves
Other Protected Areas
SCI
SPA
Tab. I. Overlapping between the various types of Protected
Areas, SCI and SPA. The table shows hectares covered by two
types at the same time.
Dolomiti Bellunesi National Park (Veneto – Northern Italy)
2.2 Physical and geographic characterization
The environmental and human presence characteri-
zations within the PA were analysed by examining
the three cartographic strata available: land cover
(CORINE Land Cover), Digital Terrain Model (DTM,
showing elevation) and road network.
The third level of the CORINE Land Cover classifica-
tion has a total of 44 classes, but in order to give a
simpler, clearer overall picture the 11 categories of
the second level were used, plus one category at the
first level including all the artificial areas mapped
3
.
An analysis of the internal composition of the PA
shows a preference for the categories “Wetlands”,
“Forest”, “Shrub and/or herba-
ceous vegetation” and “Open
Spaces” (i.e. these classes are
present in percentages exceed-
ing the national average), where-
as the remaining categories are
rarely to be found, particularly
“Artificial Areas”, “Arable Land”
and “Permanent Crops” (Fig. 14,
Table II)
4
.
The same trend, with regard to
the distribution of the CORINE
classes, is found in the individual
types of PA, although in some of
them, such as the National Parks,
“Forest”, “Shrub and/or herba-
ceous vegetation” and “Open
Spaces” are more frequent,
while “Inland Waters” and
“Wetlands” are more rarely
found.
Physical and geographic characterization
14
2.2
3
The CORINE Land Cover classes Continuous urban fabric, Discontinuous urban fabric, Industrial or commercial units, Road and rail net-
works and associated land, Port areas, Airports, Mineral extraction sites, Dump sites, Construction sites, Green urban areas, Sport and
leisure facilities have been grouped together in the class “Artificial Areas”; the classes Non-irrigated arable land, Permanently irrigated land,
Rice fields have been grouped together in the class “Arable Land”; the classes Vineyards, Fruit trees and berry plantations, Olive groves
make up the class “Permanent Crops”; the class “Pastures” stays the same; the classes Annual crops associated with permanent crops, Com-
plex cultivation patterns, Land principally occupied by agriculture with significant areas of natural vegetation, Agro-forestry areas have been
grouped together in the class “Heterogeneous agricultural areas”; the classes Broad-leaved forest, Coniferous forest, Mixed forest have been
grouped together in the class “Forest”; the classes Natural grassland, Moors and heathland, Sclerophyllous vegetation, Transitional wood-
land scrub have been grouped together in the class “Shrub and/or Herbaceous Vegetation”; the classes Beaches, dunes and sand plains,
Bare rock, Sparsely vegetated areas, Burnt areas, Glaciers and perpetual snow have been grouped together in the class “Open spaces”; the
categories Inland marshes, Peat bogs have been grouped together in the class “Inland Wetlands”; the categories Salt marshes, Salines, Inter-
tidal flats have been grouped together in the class “Coastal Wetlands”; the categories Water courses, Water bodies have been grouped to-
gether in the class “Inland waters”; the classes Coastal lagoons, Estuaries, Sea and ocean have been grouped together in the class “Marine
Waters”.
4
The internal composition of the protected areas, as regards both individual types and the system of PA, has been compared with the aver-
age situation in the Italian peninsula; this means that, at a regional level, there may be discrepancies, and even important ones, with respect
to the situation documented at a national level, and it is therefore important to emphasize that the analyses carried out at a regional level
must be evaluated in the light of the average situation pertaining in the region in question.
Fig. 14. Percentage of the various CORINE Land Cover categories in each type of
Protected Areas, in all the Protected Areas and in unprotected territory.
2.2 Physical and geographic characterization
Analyses at a regional level naturally produce a very
varied situation, with characteristics that in some
cases differ markedly from the global picture.
Abruzzi, Basilicata, Calabria and Friuli follow the
national trend fairly closely, and so do the Marche
and Apulia, although here the category “Pastures” is
also common. The case of Latium is interesting, too:
the region follows the general trend fairly well and
seems to offer particular protection (in the form of
National Reserves) to the CORINE category
“Coastal Wetlands”, which is totally protected. It is
worth noting the anomalous case of Sardinia, for
which the most important class found in the PA is
undoubtedly “Open Spaces”, while all other classes
are almost entirely unprotected, and that of Molise,
where the only type of PA to play a significant role
seems to be National Parks, concentrating chiefly on
“Forest” and “Open Spaces”.
Although this tendency towards classes involving a
more “natural” use of the soil indicates a preference
determined by strongly naturalistic criteria, it
remains to be seen whether the system is congruous
in covering significant portions of landscapes with
“widespread natural characteristics”, where the
mosaic of areas with different
types of land use produces the
well-defined, highly-prized fea-
tures of the Italian landscape.
If we compare the elevation of
the PA with the overall situation
in Italy, it can be seen that, in
general, the elevation of most of
the PA (median = 902 meters;
mean = 1,017 meters) is greater
than the mean value for Italy as a
whole (median = 337 meters;
mean = 535 meters) (Fig. 15);
this is particularly obvious in the
case of the National Parks (medi-
an = 1,043 meters; mean =
1,157 meters) and Regional
Parks (median = 971 meters;
mean = 1,054 meters). The
National Reserves (median =
763 meters; mean = 783
meters), despite having a mean
elevation greater than that of
Italy as a whole, cover a wide range of altitudes,
which therefore also include the mean values for
Italy. The Regional Reserves and Other Protected
Natural Areas, on the other hand, reflect the overall
situation in the country.
On a regional scale, elevation can of course be very
different from the situation described above. For
example, in Latium, the National Reserves and Other
Protected Natural Areas have an elevation pattern
lower than the regional average; in Tuscany, all
types of PA have an elevation pattern that perfectly
coincides with the regional average; lastly, in
Sardinia, all types, except for Other Protected
Natural Areas, have an elevation pattern lower than
the regional average.
In order to quantify human presence and to deter-
mine its influence in the natural processes that take
place in the protected areas, a comparison was
made between the presence of roads in the PA and
the average for Italy as a whole. The existence of
roads was taken as a good indication of the pres-
ence of human activity in a given area, since they
represent permanent infrastructures which, subdivid-
15
95,89 4,11 0,45 3,18 0,11 0,30 0,10 2,25 0,89
95,82 4,18 0,46 2,74 0,20 0,43 0,38 3,85 1,77
95,79 4,21 1,76 1,69 0,07 0,49 0,20 3,92 1,47
89,13 10,87 4,88 4,37 0,32 0,47 0,90 9,41 2,32
96,19 3,81 1,59 1,58 0,07 0,41 0,20 5,94 1,66
82,60 17,40 7,57 7,90 0,83 1,08 0,53 21,62 9,03
81,62 18,38 8,45 8,42 0,49 1,07 0,35 26,57 12,24
60,90 39,10 14,98 22,91 1,04 0,86 0,24 45,32 23,07
42,20 57,80 4,69 32,78 0,96 24,87 0,86 80,64 48,50
71,78 28,22 0,98 12,11 12,79 4,58 0,00 84,91 32,65
74,91 25,09 6,06 16,37 0,88 1,89 0,77 30,11 19,85
86,67 13,33 0,00 9,93 0,85 3,48 0,00 88,04 41,33
Artificial Areas
Arable Land
Permanent Crops
Pastures
Heterogeneous Agricultural Areas
Forest
Shrub and/or Herbaceous Vegetation
Open Spaces
Inland Wetlands
Coastal Wetlands
Inland Waters
Marine Waters
Unprotected Territory
Protected Areas
National Parks
Regional Parks
National Reserves
Regional Reserves
Other Protected Areas
SCI
SPA
Tab. II. Percentage of the area occupied by the CORINE Land Cover categories that falls
in each type of Protected Areas, in all the Protected Areas and in unprotected territory.
N.B.: the various types of PA, like the SCI and SPA, overlap with one another; as a result,
the percentages indicated in this table do not necessarily add up to 100%.
2.2 Physical and geographic characterization
ed into the various kinds (Motorways, Main Roads,
Secondary Roads) give a good idea of the type and
extent of human interference.
In Italy as a whole, roads cover a total of 2,199,514
ha, i.e. more than 7% of the area of the country. In
the PA, taken together, the area occupied by roads
amounts to less than 4%, indi-
cating a preference for less dis-
turbed areas. Most of this per-
centage consists of Secondary
Roads, while only a very small
area is affected by the presence
of Motorways (Fig. 16, Tab. III).
As far as the individual types
are concerned, the case of the
National Parks is particularly
interesting: here roads account
for just over 3% of the total area
and practically none of them
belong to the classes Motorways
and Main Roads. Only 687 ha
(0.05% of the area covered by
the National Parks) are affected
by the presence of Motorways,
while only 980 ha (0.09% of the area) are covered
by Main Roads.
In the case of roads, too, as described above for
CORINE Land Cover and elevation, analysis at a
regional level shows widely varying trends, with cer-
tain types of PA distinguished by the almost total
absence of roads and others by
a road density higher than the
regional average. The latter situ-
ation is to be found in Basilicata
(with the Regional Reserves), in
Campania (with the National
Reserves), in Emilia Romagna
(with Other Protected Areas)
and in the Autonomous Province
of Trento (with Other Protected
Areas). The situation found in
Aosta Valley and Friuli Venezia
Giulia is exactly the opposite:
here practically all types have a
road density approaching zero.
There are a few special cases:
Apulia, where 17% of the Other
Protected Areas are covered by
roads; Liguria, where 20% of
Other Protected Areas are cov-
ered by roads; and Emilia
Romagna, where, again, 10%
16
Fig. 15. Elevation distribution of Protected Areas, SCI and SPA. The black horizontal line
in each box indicates median altitude. The lower and upper edges of the box respectively
indicate 25% and 75% distribution percentiles. The lines outside the box indicate 95% of
elevation distribution. NP = National Parks; RP = Regional Parks; NR = National Reserves;
RR = Regional Reserves; OPA = Other Protected Natural Areas; SCI = Sites of European
Community Importance; SPA = Special Protection Areas.
Fig. 16. Percentage distribution of roads in the country, in all the Protected Areas, in each
type of PA, in the SCI and in the SPA. The percentage of territory occupied by the road
network is indicated in grey.
2.2 Physical and geographic characterization
of Other Protected Areas are covered by roads. In
these cases it must be underlined that, in view of the
very small size of the areas considered, the presence
of even a few roads can have a marked effect on
density values.
17
92,72 96,12 96,65 95,78 96,40 96,43 95,57 97,00 97,13
0,27 0,10 0,05 0,12 0,21 0,17 0,08 0,06 0,07
0,64 0,29 0,09 0,40 0,43 0,44 0,55 0,21 0,21
6,36 3,49 3,20 3,70 2,96 2,96 3,80 2,73 2,59
No Roads
Motorways
Main Roads
Secondary Roads
Italy
Protected Areas
National Parks
Regional Parks
National Reserves
Regional Reserves
Other Protected Areas
SCI
SPA
Tab. III. Percentages of roads in the country, in the Protected
Areas, in each type of Protected Areas, in the SCI and in the
SPA.
Little Egret (Egretta garzetta)
3 Nature 2000 Network in Italy
Nature 2000 Network, as described in Directive
92/43/CEE issued in 1992, is intended to be a series of
areas with the function of protecting biodiversity in the ter-
ritory of the European Community. When the network has
been completed, through agreements to be reached
between the European Union and the member countries,
it will consist of two kinds of areas: SCI (Sites of European
Community Importance) and SPA (Special Protection
Areas). Both types are currently undergoing final verifi-
cation by the European Commission, and if they are
approved they will represent the potential future for the
growth of the PA system in Italy. For the moment, the sta-
tus of these areas is that of proposed sites (SCI) subject to
transitory provisions.
In Italy, Nature 2000 Network currently consists of 343
SPA and 2,417 SCI (see Figs. 9-10, 17-18).
Taken together, the SCI cover 4,172,447 ha, corre-
sponding to over 13.8% of the area of the country. The
SPA cover 1,845,619 ha, amounting to 6% of the area
of the country.
The SCI have a mean size of 1,789 ha and a median
size of 500 ha. Only one area, out of the total of 2,417,
has an area of over 100,000 ha, while most of them
(65%) cover an area of less than 1,000 ha and 27% less
than 100 ha (Fig. 19). In the light of these figures, their
role within the protected areas system
in Italy must be given serious consider-
ation.
As far as the SPA are concerned, the
mean area is 5,381 ha, while the
median area is 1,138 ha. The largest
SPA covers an area of over 100,000
ha, while the smallest covers just over 4
ha. For both SCI and SPA, the largest
area covered is that of the Murgia Alta
district in Apulia (Fig. 19).
At a regional level, Sicily is the region
where the largest number of SCI have
been proposed (214), while Latium has
the greatest number of SPA (48) (Tab.
IV). Looking at the percentage of terri-
tory proposed for Nature 2000
Network, Liguria has identified over
25% of the region as SCI, while
Abruzzo has indicated over 29% of its area as SPA.
The environmental and physiognomic characterization of
the SCI and SPA shows that for both of them the preferred
CORINE Land Cover classes (Fig. 20, Tab. II) are the
same as for the PA, but with the addition of “Wetlands”,
“Inland Wetlands” and “Marine Waters”. Here, too,
regional characterization reveals situations that differ
considerably. In Campania, Emilia Romagna, Friuli
Nature 2000 Network in Italy
3
18
Fig. 17. Percentage of the country covered by SCI and SPA;
the number of SCI and SPA in Italy is indicated.
Fig. 18. Percentage of regional territory covered by SCI and SPA.
3 Nature 2000 Network in Italy
Venezia Giulia and Latium, the preferred classes for both
SCI and SPA include “Pastures” in addition to the above-
mentioned categories, while Abruzzi and Campania
closely follow the regional trend. A very interesting situa-
tion is to be found in certain regions, such as the Aosta
Valley, the Autonomous Province of Trento and Molise,
where both SCI and SPA seem to focus on certain
CORINE categories (essentially “Wetlands” and “Inland
Waters”) that are hardly covered at all by the PA.
On average, the SCI and SPA are to be found at eleva-
tions higher than the mean values for Italy, and with a few
exceptions the same is true for the analyses carried out at
a regional level (Fig. 15). For SPA, Sardinia, Emilia
Romagna and Tuscany have an unusual elevation distrib-
ution, below the mean values for the region.
With regard to the presence of roads, too, both SCI and
SPA roughly follow the trend of the PA and usually avoid
zones with a high road density, although the situations to
be found in the individual regions undoubtedly vary
widely, ranging from the case of the Aosta Valley, where
there are very few roads in either SCI or SPA, to that of
Molise, where the road density is very close to that of the
region as a whole.
If we compare the location of the SCI and SPA with that
of the existing PA, we find that no less than 68% of the
area of the SPA and 41% of the area of the SCI fall with-
in existing PA (Tab. I, Figs. 21-22). The biggest overlap
for the SCI is with respect to Regional Parks (21% of the
area of the SCI) and National Parks (15% of the area of
the SCI), whereas the SPA overlap with National Parks
(35% of the area of the SPA) and Regional Parks (29% of
the area of the SPA). Overlapping with other types of PA
occurs very rarely (around 2% for SCI and 5% for SPA),
although it often involves existing PA in their entirety.
There is also considerable overlapping between SCI and
SPA, involving a substantial portion of the area covered
by the SCI (33%) and most of the area covered by the
SPA (75%).
19
256.022 318.749 130 4
60.998 34.237 69 17
66.983 27.339 171 4
342.446 141.726 154 13
181.782 97.960 105 41
122.338 81.098 61 7
126.033 235.619 177 48
136.024 19.599 94 7
204.752 64.173 177 7
99.680 1.017 93 3
69.457 817 46 3
256.565 95.811 130 41
139.536 116.292 34 16
150.964 13.274 146 14
406.036 213.592 75 16
405.623 16.118 114 9
368.993 126.143 214 47
254.444 46.267 123 30
98.156 47.151 102 7
68.699 71.130 26 1
356.916 77.507 150 18
SCI (ha) SPA (ha) SCI (N°) SPA (N°)
Abruzzo
Basilicata
Calabria
Campania
Emilia Romagna
Friuli Venezia Giulia
Lazio
Liguria
Lombardia
Marche
Molise
Piemonte
Prov. Auton. Bolzano
Prov. Auton. Trento
Puglia
Sardegna
Sicilia
Toscana
Umbria
Valle d’Aosta
Veneto
Fig. 19. Number of SCI (in orange) and SPA (in blue) in Italy
by classes of area (ha).
Tab. IV. Number of SCI and SPA in each region and area (ha)
of regional territory occupied by them.
Fig. 20. Percentage of the various CORINE Land Cover cate-
gories in the SCI and SPA.
3 Nature 2000 Network in Italy
20
Fig. 21. Areas of overlapping between SCI and Protected Areas.
Protected Areas
SCI
Overlapping Areas
3 Nature 2000 Network in Italy
21
Fig. 22. Areas of overlapping between SPA and Protected Areas.
Protected Areas
SPA
Overlapping Areas
In recent years, the concept of ecological network has
come into use in many disciplines, as a reference for
both theoretical and practical applications. This wide-
spread diffusion is due to the fact that it is an extreme-
ly versatile conceptual tool, which is applicable in a
large number of contexts and which allows effective
classification of different natural and human phenome-
na, examining them as a series of factors with different
functions that intersect and cross like the mesh of a net.
We can identify four main spheres in which the con-
cept of ecological network has been applied: in land
use projects, where the network is the tool that makes
it possible to represent the dynamism and interdepen-
dence of the natural and human components; in pro-
grammes for “sustainable” socio-economic develop-
ment, where the network is used to illustrate, in a ver-
satile fashion, resources, information flows, responsi-
bilities and services compatible with the conservation
of natural resources in the area under consideration; in
designing an integrated system of protected areas and
in assessing their effectiveness; in the scientific disci-
plines of ecology and conservation biology, where the
concept of network effectively brings together the
mechanisms that determine the pattern of the various
life forms within the territory (Reggiani et al., 2000).
In the concept of network linked more closely to the dis-
ciplines of ecology and conservation biology, and in
applications relating to planning and managing land
use in particular, reference has been made to the need
to identify (and preserve) core areas in which a given
species has a stable presence, to surround these areas
with buffer zones in order to protect them from outside
influences that might be harmful, and to identify (and
preserve) landscape features – whether continuous
(corridors) or discontinuous (stepping stones) - that
enable individuals of a certain species to move
between core areas. In addition, consideration has
been given not only to the relationships between the
various components of the network, but also to those
between the said components and the environmental
matrix (Dunning et al., 1992).
Circeo National Park (Latium – Central Italy)
4 The Ecological Network: a paradigm of conceptual reference
The Ecological Network: a paradigm of conceptual reference
5
4
22
5
The points made in Chapter 4 make reference to: Boitani et al., 2002.
Crested Porcupines (Hystrix cristata)
4.1 Methodological approaches
The methodological approaches utilized to identify an
ecological network are closely connected with the atti-
tudes of the disciplines in which this concept has been
applied.
In the field of landscape ecology in particular, “net-
work” often has a strictly territorial connotation. In this
approach, the territory is interpreted and analysed on
the scale of a landscape, the degree of fragmentation
and the connectivity of its various components being
assessed as a whole.
In the case of conservation biology, the starting point
may still be a structural approach, aimed firstly at
assessing the territory and identifying its components
(Battisti, 2002), but this phase must necessarily be fol-
lowed by a phase in which the network is qualified as
to its contents by adopting the ecological perspective
of a species or group of species with respect to the ter-
ritorial system analysed. This concept of network, there-
fore, always incorporates the ecological and ethologi-
cal perspective of the species in question (Gustafson &
Gardener, 1996): the importance of a certain type of
habitat, the possible presence of a barrier or ecotone
(Manson et al., 1999) or the permeability of an envi-
ronmental matrix always refers to the particular species
under consideration.
The choice of the species is clearly a crucial point, for
which various criteria have been suggested: the con-
servationist criterion, in which the network centres on a
species or group of species that is endangered as the
result of a complex web of human and natural factors;
the bio-geographic criterion, in which the network
focuses on a species or group of species with a par-
ticularly significant distribution pattern, and lastly the
ecological criterion, in which the species included in
the network can have a key role in signalling the eco-
logical needs of other species (umbrella species), or in
highlighting the functionality of an ecosystem (key
species), or in underlining, from an ecological view-
point, the problem of territorial fragmentation (species
sensitive to fragmentation), or in providing a frame-
work for possible expansion (introduced species) etc.
(Boitani, 2000).
It is therefore clear that the choice of a species or
group of species is effective only in responding to one
particular study objective, and it is unlikely that the
results obtained can be extended to all biodiversity.
In the scientific world, this problem is at the centre of a
heated debate. On the one hand, some people are
studying the possibility, in various ecosystems, of utiliz-
ing a limited number of species to indicate the status of
total biodiversity (Oliver et al., 1998; Dobson et al.,
1997), while on the other hand this is frequently found
to be impossible (Kerr, 1997; Williams et al., 1996).
It is clear that the scientific world is actively engaged
in attempting to map the complexity of biodiversity by
means of a part of it; if it were to succeed, the reper-
cussions from the management point of view would be
far-reaching, and such attempts are therefore worthy of
great attention and participation by all the scientific
community.
In view of the fact that the scientific world has not
reached a consensus on the ethical and scientific legit-
imacy of choosing a group of species in order to eval-
uate overall biodiversity, the innovative contribution of
REN has been to broaden the foundation of the analy-
sis procedure by adopting and refining the ecological
perspective of all Italian terrestrial vertebrates and
freshwater fish.
In this context, then, the REN project consists of five
main phases of research and analysis:
1) synthesis of information on the distribution and ecol-
ogy of Italian vertebrates, leading to the creation of the
Banca Dati Faunistica (Animal Database) in 2002;
2) elaboration and analysis of the potential distribution
of each species over the national territory, through the
creation of habitat suitability models;
3) analysis and critical interpretation of each habitat
suitability model, paying particular attention to the
fragmentation of suitable areas, and proposals for the
conservation and management of suitable habitats for
each species considered;
4) validation of habitat suitability models by means of
an independent data set;
5) definition and construction of the National
Ecological Network for vertebrates.
Methodological approaches
23
4.1
The Animal Database 2002 (Banca Dati Faunistica
2002) was created as a tool necessary for the
National Ecological Network and useful for the con-
servation of all Italian vertebrates. The collection and
organization of scientific data, and the provision of
access to them, represent essential steps in any con-
servation policy.
The Animal Database 2002 consists of a complete
revision of the previous Animal Database created for
the Ministry of Environment in 1992. The revised
database includes all species of Italian vertebrates
codified in the most recent check-lists published for
the various groups (Kottelat, 1997; Bianco, 1998;
Societas Herpetologica Italica, 1996; Brichetti &
Massa, 1998; Amori et al., 1999), giving a total of
504 species. Of these, 82 species are freshwater
fish, 34 amphibians, 43 reptiles, 244 birds regular-
ly breeding in Italy and 102 mammals.
The information contained in the Animal Database
was approved by experts on the various taxa who
were called in to act as guar-
antors, ensuring that the
information was concise and
up to date.
The database is an interac-
tive system allowing consul-
tation of and access to the
information by means of the
software Microsoft Access
97© (Fig. 23). In order to
make it simple and easy to
consult, the information has
been organized as a system
of data sheets, one for each
species. Each data sheet con-
sists of four sections: a sys-
tematic taxonomic descrip-
tion, a summary of the main
biological and ecological
characteristics of the species,
an assessment of the environ-
mental features it prefers, an
updated map of its distribu-
tion range in Italy.
4.2 Animal Database
Animal Database
24
4.2
Fig. 23. Opening page of the Animal Database 2002 which includes, for each species of
Italian vertebrate, a taxonomic and systematic description (Level 1: SPECIES DETAILS), and syn-
thesis of the principal biological and ecological characteristics (Level 2: HABITAT REQUIRE-
MENTS), an assessment of habitat preferences (Level 3: HABITATS INFLUENCE) and an illustra-
tion of the distribution range in Italy (RANGE).
Habitat suitability models make it possible to develop and
summarize species-environment relationships and repre-
sent a valid tool for use in surveys and projects relating
to conservation and land use (Dupré, 1996). They pro-
vide a map showing how the various areas can offer dif-
ferent types of habitat for each species.
Through a series of procedures (Boitani et al., 2002), the
information contained in the database for each species of
Italian vertebrates was developed and translated into an
habitat suitability map covering the whole of Italy. The
models were created taking advantage of the potential of
the Geographic Information System (GIS) and utilizing
geographic data of various kinds: CORINE Land Cover,
Digital Terrain Model, water and road networks. For each
model, four suitability classes were identified (Box 1).
Habitat suitability models were drawn up for 477 of the
504 species included in the database. The species for
which there was too little information available were omit-
ted. In order to ensure that the model suited the natural
history characteristics of the species, nine different types
of model were developed.
According to the quality and quantity of data available,
the results given by the models ranged from offering no
improvement with respect to simple information about the
distribution range (i.e. the model was unable to make
any useful distinction within the territory inhabited by the
species) to identifying possible mosaics of suitable habi-
tat for the species (Fig. 24).
To give a better interpretation
of the environmental mosaics,
an analysis was then carried
out on the fragmentation of
areas with varying suitability,
for the portion of the model
included in the distribution
range only. This analysis made
it possible to study in greater
detail how the network of areas
of differing importance to
the species is structured
(McGarigal & Marks, 1999).
Each model is introduced by
notes on the distribution of
areas with different levels of
suitability, the degree of frag-
mentation and the overall per-
formance of the model (Box 2).
4.3 Environmental suitability models for the various species
Environmental suitability models for the various species
25
4.3
Suitability classes
UNSUITABLE
Environments that do not satisfy the ecological requirements of the
species.
NOT VERY SUITABLE (LOW SUITABILITY)
Habitats that can sustain the presence of the species only in a way
that does not guarantee long-term stability.
FAIRLY SUITABLE (MEDIUM SUITABILITY)
Habitats able to sustain a stable presence of the species but which do
not generally appear to be optimal habitats.
VERY SUITABLE (HIGH SUITABILITY)
Ideal habitats for the presence of the species.
BOX 1
4.3 Environmental suitability models for the various species
26
Fig. 24. Habitat suitability model for the Dunnock (Prunella modularis); the distribution range of the species includes mountainous
areas in the Alps and Apennines, while the model is able to show more clearly the areas within the range that have different degrees
of importance for the species.
Unsuitable
Low suitability
Medium suitability
High suitability
Range
Lakes and Lagoons
Habitat suitability
4.3 Environmental suitability models for the various species
27
DUNNOCK (Prunella modularis)
In Italy this species has a continuous breeding range occupying the Alpine and pre-
Alpine sectors, from the western Maritime Alps to the Carnic Alps, and also the
Apennine ridge from the area near Pavia to Molise. An isolated reproductive nucleus is
present in Calabria, on Mount Sila.
The Dunnock prefers mountainous environments, particularly forests interspersed with open spaces. In the model
(MOD1)
1
coniferous forest and moors and heathland have been considered the environmental categories most
suitable for breeding, but broad-leaved forest, mixed forest, sparsely vegetated areas (steppa, tundra) and tran-
sitional woodland shrub have also been judged fairly suitable.
Over half the distribution range (almost 60%) consists of suitable areas: 40% fairly suitable areas and 20% very
suitable areas.
The fairly suitable areas lie chiefly at the foot of the Alpine arc and along the Apennine ridge; the very suitable
areas are most frequent in the eastern Alps, while the only significant nuclei in central-southern Italy are in the
Abruzzi Apennines and on Mount Sila.
There is clearly a close link between the pattern of suitable areas
and that of the distribution range throughout the length of the
peninsula, while the fairly suitable areas, located between
Campania and Basilicata and in southern Calabria, are the only
ones of any size that seem not to be utilized by the species.
Suitable patches have been aggregated (PLADJ
2
has a high
value, equal to 84.43) and on average cover a good area
(AREA_MN
2
, corresponding to 50,10 Km
2
), even if the largest patch covers no more than 15% of the whole
range (see LPI
2
). The fairly suitable areas have a significant weight for the purposes of the continuity of suitable
habitats within the range, because if we exclude them from the analysis we see a considerable increase in the
number of patches (see NP
2
) at the expense of their area (see AREA_MN
2
, LPI
2
).
The model indicates the need to take appropriate measures to ensure the maintenance of the complex mosaic
of areas of an ecotonal kind frequented by the species in the mountainous sectors of the peninsula.
BOX 2
Suitability Classes
Unsuitable (0)
Suitable (1,2,3)
Highly suitable (3)
2.810
610
1.715
7,07
50,10
5,79
38,95
436,18
81,38
1,59
14,81
6,33
4,94
17,51
11,47
1,42
1,55
1,40
75,81
84,43
69,49
Number of
patches
(NP)
Mean patch
Size
(AREA_MN)
Km
Patch size
SD
(AREA_SD)
Largest
Patch
Index
(LPI)
Mean Shape
Index
(SHAPE_MN)
Area-Weighted
Mean
Shape Index
(SHAPE_AMN)
Percentage
of Like
Adjacencies
(PLADJ)
2
2
2
2
2
2
2
2
VALIDATION
The model for the species has been validated, with an agreement index
of 62%.
The model that includes wintering phenology
3
has also been validated
with an agreement index of 62%.
1
MOD1: this is one of the nine types of model created (for further details see Boitani et al., 2002).
2
Fragmentation indices utilized (for further details see Boitani et al., 2002).
3
For certain species of bird a further suitability model has been constructed, which also takes into consideration information relating
to phenologies other than breeding phenology (for further details see Boitani et al., 2002).
Example of description accompanying each environmental suitability model: includes notes on distribution
range, fragmentation, priority conservation activity, model performance.
Suitability Classes
Surface (km )
Unsuitable
Low suitability
Medium suitability
High suitability
Total
%
20.292,74
0,00
20.022,17
10.051,33
50.366,24
40,29
0,00
39,75
19,96
100,00
2
4.4 Validation of suitability models
The validation exercise represents a crucial moment in
the construction of a model. Since this is a conceptual
projection, and as such cannot be right or wrong, val-
idation consists in evaluating how well or otherwise the
model corresponds to the modelled phenomenon.
Consequently, this phase involves assessing the model
on the basis of how closely the picture given corre-
sponds to reality. Many different criteria can be adopt-
ed for the assessment, some more stringent than others
but all equally applicable. It is necessary, however, to
establish a basic assessment criterion, in order to dis-
criminate between models with different degrees of
correspondence to reality.
In the project, the validation exercise was carried out
by comparing the potential distribution pattern of a
given species, defined as the set of suitable areas
resulting from the model, with independent data relat-
ing to the actual presence of the species in question.
The basic criterion for assessment and the analysis pro-
cedure were chosen according to the quality of the
data available for each taxonomic group.
The validation exercise was carried out on the models
for species for which at least 10 presence location
data were available. It therefore covered 54% of the
models for mammals, 60% of those for birds, 67% of
those for reptiles, 82% of those for amphibians and
47% of those for fish. The models for fish, amphibians
and reptiles were all found to be validated, those for
birds in 68% of the cases and those for mammals in
96% of the cases.
In interpreting the results it is important to remember
that the presence location data were collected for rea-
sons unrelated to this validation exercise; as a result,
they do not always have all the requisites for a fully reli-
able assessment of the models. Nevertheless, the excel-
lent results obtained (83% of the models analysed were
found to be validated) demonstrate the predictive
capacity of the proposed models and confirm their
value as effective tools for land use management at a
national level.
Validation of suitability models
28
4.4
Arcipelago Toscano National Park (Tuscany – Central Italy)
Red Heron (Ardea purpurea)
4.5 Ecological networks
Within the framework of the project, various ecolog-
ical networks were drawn up: a “global network”,
taking into consideration all vertebrates (Fig. 25c), a
network for each taxonomic group (mammals, birds,
reptiles, amphibians and fish) (Figs. 26c, 27c, 28c,
29c, 30c) and one for the 149 endangered species
listed in the Italian Red List Book (Bulgarini et al.,
1998) (Fig. 31c). Each ecological network repre-
sents species richness distribution throughout the
country and is created by overlaying the habitat suit-
ability models for the species considered. From each
model, all the suitable zones within the distribution
range of the species have been extracted, regardless
of the level of suitability. No distinction has been
made between the various levels of habitat suitabili-
ty, so as to include all the areas with differing poten-
tial for hosting the individual species and to temper
the different ecological significance of the various
suitability classes with regard to the five taxonomic
groups.
In drawing up the ecological networks, suitability
models relating to both native and naturalized
species were taken into consideration
6
. As far as fish
are concerned, the study included non-native species
which have become acclimatized in a stable fashion
and which do not interfere, or have interfered only
slightly, with the native communities (Bianco, pers.
com.). Only validated models and those for which
there were insufficient data for the validation exer-
cise were included in the network.
For obvious reasons, the species richness values
have a different distribution pattern, as to both lati-
tude and elevation, for each ecological network
(Boitani et al., 2002).
For each network, a Biodiversity Index was calculat-
ed, consisting simply in the ratio: (Max no. of
species observed in the network)/(No. of species
present in the cell considered)*1000. The
Biodiversity Index was used both to graphically pro-
ject the ecological networks and to analyse the bio-
diversity pattern of vertebrates in Italy.
Ecological networks
29
4.5
6
Native species: species living in a given area which originated there or arrived there without the direct intervention of man. Naturalized
species: species which have been part of the Italian fauna from historic times, consisting of self-sustaining breeding groups (Andreotti et al.,
2001).
Griffon Vulture (Gyps fulvus)
4.6 Comparison between networks
One of the most interesting analyses carried out was a
comparison between the qualitative, quantitative and
geo-referenced characteristics of the various ecological
networks projected. The comparison of the overall net-
work of all species with the networks of the various
taxa and with that of endangered species demonstrat-
ed how the latter are representative for identifying the
areas of Italy that have the greatest species richness.
The degree of overlapping between the overall net-
work and the networks for the individual taxa is affect-
ed by the number of species and above all by the eco-
logical importance of the considered taxon. While
birds and mammals are amply diversified, in these
terms, this is far less true of amphibians and reptiles. It
was therefore expected that the bird network would be
the most representative and that the networks for
amphibians and reptiles would be at the opposite end
of the scale.
A comparison between the overall network and the net-
work of endangered species shows that the distribution
of the areas with the greatest species richness has a
very similar pattern in both networks; this means that
the network of endangered species can be considered
representative (an indicator) of the diversity of Italian
vertebrates.
Comparison between networks
30
4.6
Stelvio National Park (Lombardy – Northern Italy)
5 Congruency analysis of Protected Areas
The pattern of the biodiversity index for vertebrates in
Italy (with values ranging from 0 to 1,000) (Figs. 25a,
25b) was compared with the pattern of the same index
within the system of Protected Areas as a whole and
within that of the individual categories of PA (National
Parks, Regional Parks, National Reserves, Regional
Reserves, Other Protected Areas, SPA, and SCI), with
the aim of analysing which types of Protected Area
play a particularly important part in the conservation
of vertebrates. A further comparison, with the parts of
Italy not covered by PA, made it possible to pinpoint
any high biodiversity values not yet included in the PA.
A similar comparison was made, with the parts of Italy
omitted from the PA (Fig. 25d) and those omitted from
the total of PA plus SCI and SPA (Fig. 25e), in order to
assess their contribution to biodiversity conservation.
Lastly, a comparison between a map of biodiversity
indices and a map of PA (plus SCI and SPA) made it
possible to identify zones with maximum biodiversity
values that are still outside the system of PA. Two impor-
tant aspects of this study should be emphasized: a) the
analyses were carried out on a scale of 1:100,000
and the notes that follow are valid for this scale: it is
not technically correct to interpret the same data at a
greater level of detail; b) the whole analysis is based
on species diversity and does not take into account
another important parameter that is fundamental for
the PA policy: the protection of endemisms that are not
necessarily associated with high overall biodiversity
values.
The following paragraphs illustrate only the most sig-
nificant results of the study.
Congruency analysis of Protected Areas based on number of
vertebrate species
5
31
Dolomiti Bellunesi National Park (Veneto – Northern Italy)
Abruzzo National Park (Abruzzi – Central Italy)
5.1 All species of vertebrates
In the system of existing and proposed PA (all the PA
plus SCI and SPA), the pattern of the biodiversity index
differs from that of Italy as a whole, indicating that the
system of PA does not cover a random sample of land
but makes specific choices (Figs. 25a, 25b). In particu-
lar, portions of territory with high biodiversity seem to
be well represented, but portions with high diversity val-
ues that are nevertheless lower than maximum values
are poorly represented. Furthermore, the system
includes a considerable amount of land that has diver-
sity values equal to those most widespread in Italy and,
strangely, also includes portions of territory that have a
higher percentage of extremely low diversity values
than Italy as a whole. An aggregated result of this kind
is of little significance for practical purposes and sepa-
ration into areas of different types gives more informa-
tion (Figs. 25a, 25b). The areas that protect land with
high biodiversity in the most selective manner are the
National Parks and SCI, followed by Regional Parks,
although the latter also include many less important
areas, some of them with extremely low values. The
National Reserves show almost the same distribution
pattern as the index for Italy as a whole, apparently
indicating that they do not cover areas of particular sig-
nificance for the conservation of the diversity of Italian
vertebrates. The Regional Reserves have a similar trend
but have a more marked peak at high diversity values,
indicating a precise, targeted preference for protecting
limited areas. The category Other Protected Areas does
not seem to be very significant. The SPA are in an inter-
mediate situation: they include a high percentage of
areas with low biodiversity values alongside areas with
intermediate or high biodiversity. This pattern can easi-
ly be explained by the nature of the SPA, which are
dedicated to protecting species of birds in fulfilment of
European Directive 79/409/CEE: these species often
utilize open spaces, cultivated land and marginal areas
where the overall diversity value of vertebrates is rela-
tively low. It is therefore possible that the SPA fulfil the
function for which they were created perfectly well
although they are generally located in areas with low
overall biodiversity.
An analysis of the diversity index distribution maps (Fig.
25c) immediately reveals the fundamental role of the
Apennines, and to a lesser extent the Alps: they consti-
tute authentic backbones and ecological corridors that
are invaluable for Italian vertebrates. The diversity
found in the mountains and foothills is striking in com-
parison with that of the coastal areas and plains, where
diversity values are lower. Of the mountainous terrain,
at least three large areas are notable for having the
highest diversity values: the central Apennines, between
Molise and Abruzzi; the Ligurian Apennines and the
Maritime Alps; and the eastern Alps. It is important to
note the great territorial block with high mean diversity
values, running unbroken from the Casentine Forests to
the frontier with France, including the whole stretch of
the Apennines across Tuscany and Romagna, together
with the Ligurian Apennines. The existing PA cover
large portions of the central Apennines, but the rest of
the Apennine chain would be very sparsely covered
were it not for a close network of SCI which, in many
cases, helps to provide solutions of quasi-continuity
between existing PA and to cover important areas
devoid of other protection. The SCI network plays an
important role throughout much of the Apennine chain,
but there are a few gaps, which will be defined more
clearly and described more precisely in subsequent
studies dedicated to the various classes of vertebrates
(Figs. 25d, 25e): in the eastern Alps there are large
areas with high diversity values that still have no cover-
age whatsoever; in Liguria the system of PA shows a
marked bias towards certain types and is hugely depen-
dent on SCI and SPA; the Apennines in Tuscany and
Romagna are not yet adequately covered by a network
of PA; the inland areas of Molise have very high diver-
sity values but are still outside the system and this is all
the more worrying in view of the fundamental role of
the Molise Apennines as a link between central and
southern Italy; lastly, it is urgently necessary to find a
way of joining the Cilento and Vallo di Diano National
Park to the Pollino National Park, since the areas
between them are of great importance as far as the
diversity of vertebrates is concerned. An overall view
shows how few PA of any kind there are among the
foothills of the Alps and the northern part of the
Apennines. This is all the more serious in the light of the
role played by these portions of territory in maintaining
an ecological link between mountains and plains.
All species of vertebrates
32
5.1
5.1 All species of vertebrates
33
Fig. 25a. Biodiversity Index frequency pattern (see par. 4.5) for Vertebrates in the different types of Protected Areas and in the
country.
Fig. 25b. Biodiversity Index frequency pattern (see par. 4.5) for Vertebrates in the country, in the Protected Areas, in the SCI, in the
SPA, in the areas outside the Protected Areas, and in the areas outside the whole system of SCI, SPA and Protected Areas.
5.1 All species of vertebrates
34
Fig. 25c. Ecological Network of Italian Vertebrates (darker shades of red indicate suitability for a greater number of species, up to a
maximum of 182).
5.1 All species of vertebrates
35
Fig. 25d. Potential number of vertebrate species present in Italy, outside the Protected Areas.
Protected Areas
5.1 All species of vertebrates
36
Fig. 25e. Potential number of vertebrate species present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
5.2 Mammals
The National Parks seem to play a particularly impor-
tant role with regard to the entire class of mammals. In
the National Parks, the values of the diversity index for
this class fall entirely within the peak of maximum
diversity, reflecting one of the criteria most frequently
adopted in choosing a National Park, i.e. the presence
of species of large vertebrates. The Regional Parks, on
the other hand, despite having a peak of high diversi-
ty greater than that of Italy as a whole, also include a
greater proportion of areas with very low diversity,
indicating that the importance of the areas covered is
often unconnected with the presence of a particularly
important fauna (Figs. 26a, 26b, 26c).
The Regional Parks, taken together, cover a very wide
area, influencing the distribution trend of the whole sys-
tem of existing and proposed PA, which shows a simi-
lar diversity pattern.
National Reserves and Regional Reserves, too, include
sectors with a high diversity of mammals, but they also
include many areas with medium and low diversity.
The SCI show three frequency peaks corresponding to
high, average and low diversity values for mammals,
confirming that they have been proposed for specific
reasons, often calibrated on only a few species, and
not only mammals (Figs. 26a, 26b).
Mammals are found over a very wide range of differ-
ent ecological scales, from the microhabitats of the
smallest insectivores to the extensive areas occupied by
the largest, most mobile species, and it is certainly dif-
ficult to make any kind of generalization, but they
undoubtedly constitute one of the classes for which it is
most necessary to adopt an authentic ecological net-
work approach, in order to ensure mobility and links
between fragmented metapopulations. An effort can
probably be made to modify the boundaries and
dimensions of the SCI in order to emphasize their sup-
porting role for the system of Other Protected Areas (as
links between parks and buffer zones around their
boundaries), particularly in the case of National and
Regional Parks (Fig. 26c).
The diversity distribution maps show at least three
areas with very high mammal diversities that have not
yet been included in any protected area, either exist-
ing or proposed (Figs. 26d, 26e). The eastern Alps
(northern and eastern Friuli), despite the presence of a
good number of SCI complementing a few Other
Protected Areas, deserve particular attention with
regard to setting up a network of PA that are effective-
ly and efficiently connected to one another. Good net-
work management might be an alternative to extend-
ing the area covered by the protected areas. In west-
ern Piedmont, the foothills between Pinerolo and
Cuorgnè have been found to have extremely high
mammal diversity, and yet the area is devoid of pro-
tection: since this is a transversal area running across
several hydrographic basins, action to ensure continu-
ity within the area should be given high priority.
Lastly, the area of the mountain pass between Tuscany
and Emilia to the east of Abetone, though surrounded
by a few PA and numerous SCI, is still outside the sys-
tem, despite the fact that it is an important node in the
mammal network of the northern Apennines.
Mammals
37
Ibex (Capra ibex)
5.2
5.2 Mammals
38
Fig. 26a. Biodiversity Index frequency pattern (see par. 4.5) for Mammals in the various types of Protected Areas and in the
country.
Fig. 26b. Biodiversity Index frequency pattern (see par. 4.5) for Mammals in the country, in the Protected Areas, in the SCI, in the
SPA, in the areas outside the Protected Areas and in the areas outside the whole system of SCI, SPA and Protected Areas.
5.2 Mammals
39
Fig. 26c. Ecological network of Italian Mammals (darker shades of red indicate suitability for a greater number of species, up to a
maximum of 57).
5.2 Mammals
40
Fig. 26d. Potential number of mammal species present in Italy, outside the Protected Areas.
Protected Areas
5.2 Mammals
41
Fig. 26e. Potential number of mammal species present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
5.3 Birds
Our greater scientific knowledge of this class, together
with the greater visibility of the species and the avail-
ability of a large number of research workers, both
professionals and volunteers, has meant that birds
have been one of the most frequent reasons given for
setting up protected areas. Species diversity does not
seem to have been the most important parameter, how-
ever, but rather the presence of a few particular
species. The diversity pattern throughout Italy has a
characteristic bimodal trend with two relatively central
peaks. The system of PA, as a whole, effectively reflects
this trend, indicating a general lack of preference for
areas with higher diversity levels and indeed, on the
contrary, including a disproportionate quantity of
areas with a very low bird diversity level. The National
Parks are confirmed as the type of protected area that
most frequently includes territory with high diversity,
while the Regional Parks, National Reserves and in
general all other types of protected area do not seem
to have any specific preference for this class. The SCI,
on the other hand, are notable as being the areas with
the highest bird diversity, whereas the SPA show a less
clear-cut preference, very probably for the reasons
already explained above (Figs. 27a, 27b, 27c).
Perhaps for no class as much as that of birds, the glob-
al network, consisting of PA plus SCI and SPA, appears
to give adequate coverage to all the areas with the
highest diversity. With the exception of Molise, where
there are still large portions of territory with high diver-
sity that do not form part of the system of PA (Figs. 27d,
27e), there do not seem to be other large areas of ter-
ritory that have high diversity and yet are devoid of
protection. On the northern slopes of the Apennines in
Tuscany and Romagna there is a vast area with high
diversity that appears to be very fragmented, but a
close-set mosaic of SCI seems to form a homogeneous
network, the ecological efficiency of which, however,
should be verified.
In central-southern Piedmont there is a wide area with
average diversity that does not seem to have an ade-
quate number of protected areas: the extension of the
area and its central position, overlooking the Po plain,
make it potentially very interesting for conservation;
the territorial layout of the protection system should be
properly verified. The foothills in Piemonte, too, have
average diversity and are poorly protected: as already
mentioned, the role of this area as a link makes it more
important, despite the fact that the diversity index is not
very high.
Birds
42
5.3
Peregrine Falcon (Falco peregrinus)
Snow Finch (Montifringilla nivalis)
5.3 Birds
43
Fig. 27a. Biodiversity Index frequency pattern (see par. 4.5) for Birds in the various types of Protected Areas and in the country.
Fig. 27b. Biodiversity Index frequency pattern (see par. 4.5) for Birds in the country, in the Protected Areas, in the SCI, in the SPA,
in the areas outside the Protected Areas, and in the areas outside the whole system of SCI, SPA and Protected Areas.
5.3 Birds
44
Fig. 27c. Ecological Network of Italian Birds (darker shades of red indicate suitability for a greater number of species, up to a maxi-
mum of 86).
5.3 Birds
45
Fig. 27d. Potential number of bird species present in Italy, outside the Protected Areas.
Protected Areas
5.3 Birds
46
Fig. 27e. Potential number of bird species present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
5.4 Reptiles
The pattern of the biodiversity index for reptiles mirrors
that of amphibians and shows a single peak, corre-
sponding to the highest levels of biodiversity, indicat-
ing that, on the scale used to interpret the analyses,
reptile species are more ubiquitous and have larger
distribution ranges (Figs. 28a, 28b, 28c). In this con-
nection, the various types of protected areas show
trends for the diversity index that are essentially similar
to one another and also to the trend of Italy as a
whole. In other words, the PA do not seem to show any
significant preference for areas with greater reptile
diversity. Once again, the National Parks cover the
peak of high diversity, while the category Other
Protected Areas covers a single peak coinciding with
extremely high diversity.
It is probably better to look at the distribution maps in
order to identify any irregularity between biodiversity
distribution and the PA system. Reptile distribution most-
ly concerns the Italian peninsula, and the areas of
greatest diversity that appear to be least covered by
the network of PA are central and western Liguria and
central and southern Tuscany (Figs. 28d, 28e). In
Liguria, the network of proposed SCI works well to
cover mountainous and inland territory, but the coastal
strip still seems very unprotected. In Tuscany, most of
the hills between the hinterland of Livorno, the
Metalliferous Hills and the Maremma, as far as the
boundary with Latium, have no protection. In southern
Italy, Calabria and eastern Sicily stand out as areas
with high diversity levels, but they are only partially
included in the system of protected areas: in Calabria
it is chiefly the coastal strips that remain unprotected
and in south-eastern Sicily the SCI cover only a small
percentage of the most important areas. It should be
emphasized that the completion of the ecological net-
work for reptiles (as for certain other classes) does not
necessarily have to be achieved by setting up new pro-
tected areas; often it is sufficient to ensure habitat suit-
ability by means of rules and regulations regarding
human activities, agricultural practices and the use of
chemicals.
Reptiles
47
Grass Snake (Natrix natrix)
5.4
European Pond Turtle (Emys orbicularis)
5.4 Reptiles
48
Fig. 28a. Biodiversity Index frequency pattern (see par. 4.5) for Reptiles in the various types of Protected Areas and in the
country.
Fig. 28b. Biodiversity Index frequency pattern (see par. 4.5) for Reptiles in the country, in the Protected Areas, in the SCI, in the
SPA, in the areas outside the Protected Areas, and in the areas outside the whole system of SCI, SPA and Protected Areas.
5.4 Reptiles
49
Fig. 28c. Ecological Network of Italian Reptiles (darker shades of red indicate suitability for a greater number of species, up to a
maximum of 15).
5.4 Reptiles
50
Fig. 28d. Potential number of reptile species present in Italy, outside the Protected Areas.
Protected Areas
5.4 Reptiles
51
Fig. 28e. Potential number of reptile species present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
5.5 Amphibians
As might be expected in view of their unusual natural his-
tory and the number of species, amphibian diversity is
essentially concentrated on average-low values through-
out most of Italy (Figs. 29a, 29b, 29c). The system of PA,
on the other hand, shows a clear-cut preference for areas
of high diversity as well. The National Parks and, to a
lesser extent, the Regional Parks and the National and
Regional Reserves contribute to this preference. A key
contribution towards enabling the system of PA to safe-
guard a good proportion of the areas with high amphib-
ian diversity comes from the SCI, and still more the SPA.
The latter components of the system are chiefly located in
areas of interest for bird species that inhabit wetlands
and so they naturally provide conditions that are equally
suitable for amphibians.
It can be seen from looking at the distribution maps, on
the large scale used for the whole study, that two impor-
tant components of the ecological network for amphib-
ians are worthy of particular attention. The first is the
area of the Apennines in Tuscany and Emilia, and still
more the Ligurian Apennines as far north as Genoa,
where we find the highest levels of amphibian diversity
that are not well covered by the network of PA. The poor
mobility of the species and the fact that they are associ-
ated with ecological conditions that do not generally
pertain over extensive areas mean that the ecological
network project should be studied with particular care:
the area in question has a good number of SCI but large
portions of suitable territory still remain outside the sys-
tem and activities must therefore be carefully regulated
so as to keep possible links operating efficiently (Figs.
29d, 29e).
The second component of the ecological network for
amphibians is the massive block of areas with average
and high diversity that runs down from Molise to the
Pollino National Park. This wide strip of land has a low
density of protected areas and SCI, perhaps on account
of the absence of other important biodiversity factors.
The map shows that much of this strip of land is poten-
tially suitable for amphibians and it is to be hoped that
the SCI system will be modified to ensure that it gives
adequate protection to this class of vertebrates (Figs.
29d, 29e).
Amphibians
52
5.5
Pool Frog (Rana lessonae)
Tyrrhenian Painted Frog (Discoglossus sardus)
5.5 Amphibians
53
Fig. 29a. Biodiversity Index frequency pattern (see par. 4.5) for Amphibians in the various types of Protected Areas and in the
country.
Fig. 29b. Biodiversity Index frequency pattern (see par. 4.5) for Amphibians in the country, in the Protected Areas, in the SCI, in the
SPA, in the areas outside the Protected Areas, and in the areas outside the whole system of SCI, SPA and Protected Areas.
5.5 Amphibians
54
Fig. 29c. Ecological Network of Italian Amphibians (darker shades of red indicate suitability for a greater number of species, up to
a maximum of 13).
5.5 Amphibians
55
Fig. 29d. Potential number of amphibian species present in Italy, outside the Protected Areas.
Protected Areas
5.5 Amphibians
56
Fig. 29e. Potential number of amphibian species present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
5.6 Fish
It is difficult to examine fish using the same methods
that are found useful for other vertebrates, and the
observations that can be made for this class are so
general that more detailed analyses are necessary in
order to develop directly applicable guidelines. The
graph illustrating the pattern of biodiversity indices for
fish shows a trend which, with a few variations, covers
all the values of the index in an almost uniform fashion,
indicating that there is a wide variety of situations,
from the simplest where only a few species are present
to those with a greater number of species (Figs. 30a,
30b. 30c). In the system of PA, there seems to be a
clear preference only for areas with low diversity, while
the system fails to protect the most significant part of
the territory for this class of vertebrates. Even the
National Parks and SCI do not modify this situation,
and only one category, Other Protected Areas, shows
a peak corresponding to the highest diversity values. It
would seem clear that, apart from the difficulty of inter-
preting these figures, fish have not been given ade-
quate consideration in planning and creating protect-
ed areas in Italy, not even in the SCI.
The distribution map shows, despite the obvious frag-
mentation of the hydrographical networks, that the
areas of greatest diversity are to be found in the Po
plain and in the basins of Tuscany and Latium, which
should be the object of a renewed conservation strate-
gy (Figs. 30d, 30e): while it is possible to achieve this
objective by means of legislation and action that does
not necessarily make reference to territorial institutions
such as protected areas, it is nevertheless desirable to
develop the utmost synergy with the PA system; territo-
rial coverage should therefore be adapted where pos-
sible to take account of the diversity of fish species,
too, so as to promote their conservation.
Fish
57
Chub (Leuciscus cephalus)
5.6
5.6 Fish
58
Fig. 30a. Biodiversity Index frequency pattern (see par. 4.5) for Fish in the various types of Protected Areas and in the country.
Fig. 30b. Biodiversity Index frequency pattern (see par. 4.5) for Fish in the country, in the Protected Areas, in the SCI, in the SPA, in
the areas outside the Protected Areas, and in the areas outside the whole system of SCI, SPA and Protected Areas.
5.6 Fish
59
Fig. 30c. Ecological Network of Italian Fish (darker shades of red indicate suitability for a greater number of species, up to a maxi-
mum of 35).
5.6 Fish
60
Fig. 30d. Potential number of fish species present in Italy, outside the Protected Areas.
Protected Areas
5.6 Fish
61
Fig. 30e. Potential number of fish species present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
5.7 Endangered species
The network of endangered species (149 species) is
of critical importance for biodiversity conservation.
Special attention should be paid to these species in
drawing up the system of PA and the legislation
intended to protect them, because their vulnerable
state means that fewer options are available in
choosing protected areas and corridors between
them. For this reason, the analysis of graphs and
maps must be still further refined. The pattern of the
diversity index of endangered species shows two
peaks corresponding to low and average diversity
values, together with a less pronounced peak for
high diversity values (Figs. 31a, 31b, 31c). This
poses a problem with regard to the distribution of the
protected areas, not only in the zones with the
greatest diversity but also in those with low diversity
values, because in the case of endangered species,
under national and European legislation, no possible
means of conservation can be ignored. The system
of PA, both existing and proposed, seems to answer
the protection requirement of this group of species
relatively well and an interesting synergy can be
seen between the role played by National Parks, SCI
and SPA on the one hand and that of Regional Parks
and reserves on the other: the former are concen-
trated in the areas of greatest diversity while the lat-
ter also cover areas with low diversity indices.
Although the general structure of the system appears
to be valid, however, an examination of the distribu-
tion maps reveals incongruities of considerable
magnitude. There is confirmation of the three macro-
areas with the highest diversity values where there
are still large portions of territory outside the system
of PA (Figs. 31d, 31e): the eastern Alps, central and
western Liguria, the Apennines of central Italy. In
these priority areas, the system of PA must become a
more efficient network as far as the biological needs
of the endangered species are concerned; corridors
between the protected areas must be created and
maintained so that the system effectively constitutes a
network and not just a collection of protected areas
in the most critical places. In Friuli, it is urgently nec-
essary to maintain an efficient link between Slovenia
and the Belluno area, and not only at high altitudes.
In Liguria, it is the hinterland of Savona and Imperia
that deserves particular attention, while in the central
Apennines the critical areas are those that are not
yet protected, especially the areas north of the
Simbruini range in Latium and the corridors in
Molise linking the great National Parks to the
Regional Parks and reserves located further to the
south.
Besides these three macro-areas, we must also men-
tion a few other situations which require immediate
intervention in order to strengthen the system and
form a network (Figs. 31d, 31e): the link between
the Cilento and Vallo di Diano National Park and the
Pollino National Park can wait no longer, and the
same is true of the completion of the protection of the
coastal chain, a factor that is essential for maintain-
ing the vitality of the protected areas of Calabria,
from Mount Sila to Aspromonte. Careful observation
of this map reveals other situations, too, that could
be rectified with a minimum of effort: better protec-
tion could be given to the foothills of Lombardia and
Piedmont, the lower slopes of the Apennines in
Tuscany, the Metalliferous Hills and a few critical
fragments in the Gargano, Sicily and Sardinia.
Taking into account the limitations imposed by the
type of data utilized with regard to scale, these
guidelines seem sufficiently robust to demand careful
verification in the field, in order to identify the best
operational solutions for conservation, whether by
extending and correcting the existing/proposed sys-
tem or by introducing regulations and laws specifi-
cally covering the endangered species.
Endangered species
62
5.7
5.7 Endangered species
63
Fig. 31a. Biodiversity Index frequency pattern (see par. 4.5) for endangered species (Bulgarini et al., 1998) in the various
types of Protected Areas and in the country.
Fig. 31b. Biodiversity Index frequency pattern (see par. 4.5) for endangered species (Bulgarini et al., 1998) in the country, in the
Protected Areas, in the SCI, in the SPA, in the areas outside the Protected Areas, and in the areas outside the whole system of SCI,
SPA and Protected Areas.
5.7 Endangered species
64
Fig. 31c. Ecological Network of Italian endangered species (Bulgarini et al., 1998) (darker shades of red indicate suitability for a
greater number of species, up to a maximum of 54).
5.7 Endangered species
65
Fig. 31d. Potential number of endangered species (Bulgarini et al., 1998) present in Italy, outside the Protected Areas.
Protected Areas
5.7 Endangered species
66
Fig. 31e. Potential number of endangered species (Bulgarini et al., 1998) present in Italy, outside the Protected Areas, SCI and SPA.
Protected Areas, SCI, SPA
6 Congruency analysis of Protected Areas
The efficacy of protected areas (PA) for the protection
of vertebrate biodiversity has been evaluated in the
first part of this work through the examination of the
contents of the PA system. In particular, a presence –
absence analysis for both species and habitat types
inside and outside the protected areas system (see
Chapter 4 and 5) has been done. The results obtained
are very encouraging and underline the overall impor-
tance of the Italian PA system.
More sophisticated techniques have been used in the
second part of this work to carry out the same type of
analysis. In particular, the first application of the irre-
placeability concept to the Italian context has been test-
ed. Irreplaceability is a measure strictly linked to the
importance of an area for the conservation of natural
resources. In fact, if an area has no substitute or a
small number of substitutes in the conservation plan,
then it is characterized by high irreplaceability values.
Conversely, low irreplaceability values indicate that the
area is relatively unimportant for reaching the conser-
vation goal considered for a given region.
In this sense, irreplaceability is the measure given to an
area (terrestrial or marine) that reflects the importance
of that area in the context of the considered region,
and within the limits of the conservation target that
have been set. Briefly, the concept of irreplaceability
can be explained as the probability that the protection
of a given area is classified as necessary in order to
reach an established conservation target. This defini-
tion gives us the opportunity to use a quantitative
approach to conservation, considering the different
features that characterize each site (habitat types, suit-
ability for a species, presence or absence of a
species).
Pressey and colleagues (1994, 1995) proposed the
proportion of all the possible combinations of sites that
are able to reach the target and that contain a given
site as a measure for the irreplaceability of that site.
For example, consider the situation where the PA sys-
tem is made by selecting a combination of n sites from
a total of t sites. The number, C, of possible combina-
tions of n sites that can be drawn is calculated using
the formula:
C = t!/n!(t-n)!
The number C groups all the possible combinations,
but only a small sub-group of C meets the needs linked
to the conservation target. All the other sites do not
meet the conservation target for one or more features.
Moreover, for each site x, the possible combinations of
sites can be divided into two sets, one groups all the
combinations that include the site x while the other
groups all the combinations that do not include the site
x. According to Pressey et al. (1994), the irreplace-
ability of a site can be calculated as the ratio between
the number of combinations of sites that include site x
and the number of all the possible representative com-
binations.
Further investigations (Pressey et al. 1994, New South
Wales National Parks and Wildlife Service, 2001)
have suggested different possible improvements of the
irreplaceability measures. In fact, the set of combina-
tions that include the site x can be subdivided in two
groups: those combinations that would not be any
more representative if site x would have been exclud-
ed and those combinations that would still be repre-
sentative even if the site x would have been excluded
(i.e. those combinations for which the site x is redun-
dant). A better measure of irreplaceability can then be
calculated using the ratio between the number of com-
binations that contain the site x (and that would not be
representative if the site x would have been excluded)
and the number of all representative combinations.
However, the above definitions are difficult to be used
when calculating the irreplaceability for a regional or
national data set because of the exponential nature of
the problem: if the number of sites considered increas-
es, the number of possible combinations increases
exponentially, rapidly reaching numbers that are
impossible to manage even with the fastest and newest
supercomputers. For these reasons the irreplaceability
values are calculated using a statistical approach, in
particular using the methods described in Ferrier et al.
(2000).
All the analysis have been done using the software C-
Plan 3.20 (New South Wales National Parks and
Congruency analysis of Protected Areas based on irreplaceability
6
67
Wildlife Service, 2001), that, together with a
Geographic Information System, allows to map all pos-
sible options necessary to reach an established con-
servation target. All calculations are based on the sites
by features matrix, where the rows represent the sites
and the columns represent the environmental features.
The latter being represented by the habitat suitability
models and the distribution ranges for the Italian verte-
brates described in Chapter 4. It should be clearly
reminded that some species do not have a habitat suit-
ability model and this should be considered when inter-
preting the results obtained, especially in the case of
reptiles, amphibians, and breeding birds. In fact, for
these three groups some of the species that have no
habitat suitability model are endemic and hold a great
conservation value.
In order to calculate the irreplaceability values, C-Plan
requires the study area to be divided into planning
units (the spatial elements that constitute the basis for
all the analysis). Since there is no possibility to design
meaningful planning units, the entire national territory
has been divided into more than 13,000 square cells
(5 km by 5 km) and for each cell the areas occupied
by the distribution range of each species and by the
highly suitable territory for each species have been cal-
culated. The matrix of data (Tab. V) has been used to
estimate the irreplaceability for each planning unit and
the results has been imported into ArcGis 8.3 software
in order to produce the final maps.
C-Plan 3.20 is able to calculate irreplaceability values
also considering each species of vertebrate with a dif-
ferent weight, calculated according to its degree of vul-
nerability. Thus, a number of international and nation-
al conventions, treaties and laws, together with bio-
logical indexes (Tab. VI) have been considered for
each species in order to rank all species, according to
their degree of vulnerability.
The international and national conventions, treaties
and laws have been considered all together to give
each species a score scaled between 0 and 100
depending on the number of conventions in which the
species has been listed. The points deriving from the
conventions have been summed to all the other values
and the final values have been scaled between 0 and
100. The ranking resulting from these operations
seems to satisfactorily reflect the conservation values of
each species. In fact, species like Pelobates fuscus,
Euproctus platycephalus, Podarcis wagleriana, Vipera
ursinii, Aythya nyroca, Cervus elaphus corsicanus,
Lampetra zanandreai, Rupicapra pyrenaica ornata
resulted having the highest values, and species like
Sylvilagus floridanus, Mustela vison, Anguilla anguilla
and Cyprinus carpio having the lowest values.
Obviously, some species appear to rank with a value
higher or lower than expected. This is the case of the
6 Congruency analysis of Protected Areas
68
Tab. V. Example of matrix of data used for the analysis in C-
Plan. PUID indicate the planning units’ number, all the following
columns indicate the species of vertebrates, and the numbers in
the matrix indicate the number of hectares of highly suitable ter-
ritory in each planning unit.
Tab. VI. List of variables considered for each taxonomic class.
(1) Special protected species; (2) According to Bulgarini et al.
1998, Amori et al. 1999, Societas Herpetologica Italiana
1996, Gandolfi et al. 1991; (3) According to Brichetti &
Gariboldi 1997; (4) According to Bulgarini et al. 1998; (5)
According to Bulgarini et al. 1998; (6) According to IUCN
2001, Amori et al. 1999, Bulgarini et al. 1998; (7) According
to Tucker & Heath 1994.
6 Congruency analysis of Protected Areas
wild boar (Sus scrofa) that occupy a position in the cen-
tral lower part of the ranking because, even though it
is considered a pest in Italy and the number of animals
is extremely high, there is the possibility of an endem-
ic subspecies still surviving in some areas of the Italian
peninsula. The same kind of considerations can be
done for species like the wolf (Canis lupus) and the
peregrine falcon (Falco peregrinus) that have a very
high position in the ranking; these two species have
seen a great increase in their number in the last ten
years for many different reasons, but all the interna-
tional and national conventions/laws have not been
updated, thus both species are still considered very
endangered. Conversely, species like Lepus corsicanus
was re-discovered only very recently and so is not con-
sidered in many conservation treaties or in the IUCN
red list. This made that such species occupies a low
position in the ranking.
The list created in this way has been split into five
groups, each group containing 20 points: the first
group going from 81 to 100, the second from 61 to
80, the third from 41 to 60, the fourth from 21 to 40,
and the fifth all the remaining values. Each of the five
groups correspond to a vulnerability degree, with the
first group (species with point between 100 and 81)
being the most endangered, and the fifth group
(species with point between 0 and 20) being the least
endangered. The number of groups to consider has
been chosen according to the specifications of the soft-
ware C-Plan, which is not able to manage more than 5
vulnerability groups.
One of the most important elements for the analysis is
the conservation target that is used by C-Plan to calcu-
late the irreplaceability index. The conservation target
can be seen as the result that should be obtained for a
given region and it can be established following many
different criteria. In particular, it is possible to establish
a single target for all the elements considered in the
analysis (the species of vertebrates) or a different tar-
get for each feature (New South Wales National Parks
and Wildlife Service, 2001). The first hypothesis (a
fixed target for all species) has been chosen in this pro-
ject, because the intent of the analyses is to depict a
general vision for the Italian situation. Moreover, there
is no possibility with the current knowledge of the
Italian fauna to establish a biologically meaningful con-
servation target for each species. The general conser-
vation target has been set at 20% of the highly suitable
habitat or of the distribution range for each species,
because 20% of the Italian territory is currently inter-
ested by existing or proposed protected areas (see
Chapters 2 and 3).
All analyses previously described have been carried
out also considering the already existing protected
areas. C-Plan is able to consider the existence of pro-
tected areas in the analysis and in this way it is possi-
ble to identify all the areas that are not protected but
should be so in the future possible scenarios. To carry
out this type of analysis a planning unit has been con-
sidered as protected only if at least 50% of its area is
covered by protected areas.
All analyses considered have been carried out for all
the vertebrates together and for the single taxonomic
groups (mammals, birds, reptiles and amphibians,
freshwater fish). The values of irreplaceability have
been subdivided into 5 classes in order to facilitate the
visualization and the interpretation of the maps. In par-
ticular, the irreplaceability values have been standard-
ized (the mean has been subtracted from each value
and the difference has been divided by the standard
deviation); in this way each irreplaceability value has
been transformed in a negative number (if the original
value is less than the mean) or a positive number (if the
original value is greater than the mean) that give a
measure of the difference (in standard deviation units)
among each value and the mean value. The classes
used in the maps show an important division: the first
class (in yellow) has all values smaller or equal to 0
(that is all the values smaller than the mean), while the
four subsequent classes (shaded in red hues) have the
values among 0 and 1 (irreplaceability values greater
than the mean and smaller than the mean plus one
standard deviation), among 1 and 2 (irreplaceability
values greater than the mean plus one standard devia-
tion and smaller than the mean plus two standard devi-
ations), among 2 and 3 (irreplaceability values greater
than the mean plus 2 standard deviations and smaller
than the mean plus three standard deviations), and
greater than 3 (irreplaceability values greater than the
mean plus three standard deviations). In this way it is
possible to clearly show on a map the planning units
that more than others are important for the conserva-
tion of the vertebrates.
69
Figure 32a (that consider the distribution range for all
the vertebrates) shows the most important areas for the
Italian vertebrates. The eastern Alps are one of the
most interesting locations in the entire Italian peninsu-
la, and extremely important are also some areas of the
Padana plain, the Apennines, the Apulia region and
Sardinia. This map, however should be considered
carefully, since it is greatly influenced by the birds (that
constitute great part of the species) and by some other
species of the different groups (like some amphibians
in Sardinia, some bats in the Padana plain and some
freshwater fish in relation to some lakes). Moreover,
Fig. 32a presents the results obtained considering the
distribution range of the species of vertebrates, and so
it is probably influenced by the low precision that char-
acterize some of these ranges. The picture outlined
with this analysis is very different from that built using
just the number of species, in particular in correspon-
dence to the Apennines, which is considered as no
important even if it hosts a great number of species.
The situation outlined in Fig. 32b is completely differ-
ent. In this case the analyses have been done consid-
ering the habitat suitability models for all the verte-
brates. The results are much more detailed, and the
areas considered as important have a spatial distribu-
tion that seems to be much more meaningful from a
landscape point of view. Also in this case the eastern
Alps can be considered as one of the most important
areas for the Italian vertebrates, but in this map also
the Apennines, as should be expected considering the
high number of species present in the area, play an
important role for biodiversity conservation.
Conversely, many areas in the plains have much less
importance as compared with the Fig. 32a. Some of
the differences among the two maps can be explained
because there is not a habitat suitability model for
some high ranked species of vertebrates (like Larus
audouinii, or Speleomantes spp.), but by far the most
important difference is the fact that the distribution
range includes many areas where the species are not
really present (because those area are not suitable for
the species considered), while the habitat suitability
models represent a much more detailed sketch of the
species actual or potential distribution.
Figure 32c shows an interesting result of the analysis
carried out for all the vertebrates and also considering
the presence of protected areas. In this case, the map
outlines which areas still outside the existing protected
areas system should be considered with greater atten-
tion. Even taking into account the limitations of an
analysis that considers just the distribution ranges of
the species, the result underlines the importance of
Sardinia (that lacks of protected areas for great part of
its territory) and also of many areas of the eastern Alps
without any protection, and in many cases completely
uncovered also by SCI and SPA.
Figure 32d (obtained from habitat suitability models
for all vertebrates, and considering also the existing
protected areas) is almost the same as Fig. 32c, even
if some areas are classified differently and the most
important areas are much more fragmented then those
in Fig. 32c. However, once more it has been under-
lined the importance of the eastern Alps and of
Sardinia.
Alpine Marmot (Marmota marmota)
6.1 All species of vertebrates
All species of vertebrates
70
6.1
6.1 All species of vertebrates
71
Fig. 32a. Distribution of the irreplaceability values in Italy. Values calculated using the distribution ranges of all the Italian verte-
brates.
6.1 All species of vertebrates
72
Fig. 32b. Distribution of the irreplaceability values in Italy. Values calculated using the habitat suitability models for all the Italian
vertebrates.
6.1 All species of vertebrates
73
Fig. 32c. Distribution of the irreplaceability values in Italy. Values calculated using the distribution ranges of all the Italian vertebrates
and considering the existing Protected Areas.
6.1 All species of vertebrates
74
Fig. 32d. Distribution of the irreplaceability values in Italy. Values calculated using the habitat suitability models for all the Italian
vertebrates and considering the existing Protected Areas.
6.2 Mammals
Mammals are probably the second best known
group of Italian vertebrates after birds. In particular,
the species-habitat relationships used to build their
habitat suitability models are quite accurate, and the
data used to build their models are particularly suit-
ed to describe mammal distribution. For this reason
the results obtained for the mammals are probably of
great utility.
Figure 33a provides a first cut of the conservation
status of the Italian mammals (analysis carried out
using the distribution ranges of mammals). In this
case the most important areas are localized in cor-
respondence to the eastern Alps. Particularly impor-
tant is the boundary area among Italy, Austria and
Slovenia (corresponding at least in part with the
areas of presence of the bear) that is not covered by
any existing protected areas, and is covered only for
a very small portion by SCI. Also in the eastern Alps
there are many other important areas, and most of
them are not covered by any protected area. The sit-
uation improves greatly considering also the SCI that
cover with greater efficacy some of the areas classi-
fied as important by the analysis. In the western
alpine range, the situation is slightly different, since
there are a few protected areas that effectively cover
important areas. Once more, some of the SCI (espe-
cially in Piedmont) seem to be extremely important,
since they cover areas important for the Italian mam-
mals, but with no protection. The Apennine range
has been classified as relatively important, with
areas, like Mount Amiata and the Abruzzi moun-
tains, of great interest and relatively well covered by
the existing PA system (at least in Abruzzi). Quite dif-
ferent is the situation in Sardinia, where the areas
important for the mammals (in the southern part of
the island) are not covered by any protected areas
(except for the small Reserve of Monte Arcosu),
either existing or proposed.
Figure 33b depicts a general sketch for the conser-
vation of Italian mammals using the habitat suitabili-
ty models. The eastern Alps are still among the most
important areas in the whole peninsula, but the
areas of greatest importance are much more restrict-
ed. Again, in this case nor PA cover the most impor-
tant areas for the eastern Alps, neither SCI and SPA
are located accordingly to the important areas.
Conversely, for the Apennines the situation outlined
by Fig. 33b is completely different from that of Fig.
33a. In fact, Fig. 33b outlines that almost all
Apennines are important for the Italian mammals,
especially in the central and northern part of the
chain, but also in the southern part as well as in
some areas of Sicily. In Sardinia the eastern part of
the island is classified as important (contrary to what
outlined by the analysis done using just the distribu-
tion range), and once more the existing protected
areas do not cover almost any important area. In this
case SCI are of great importance because they cover
great part of the gaps in the protected areas system
of the region. Moreover, Fig. 33b outlines the impor-
tance of some planes, especially in Sicily, Apulia
and in the Padana plain, especially for micro-mam-
mals like Neomys fodiensis, Suncus etruscus,
Crocidura sicula, Lepus europaeus and other
rodents.
The map presented in Fig. 33c shows the results
obtained considering not only the distribution range
of the Italian mammals, but also the already existing
protected areas. The first important consideration is
that the distribution ranges for the mammals are
Mammals
75
Hedgehog (Erinaceus europaeus)
6.2
6.2 Mammals
quite general and are not able to depict a particu-
larly meaningful picture in this analysis. However, it
is possible to note that Sardinia is by far the most
important part of Italy still outside the protected
areas system.
The situation outlined in Fig. 33d is completely dif-
ferent. In this case the analysis considers the pres-
ence of the protected areas already existing and the
habitat suitability models for the mammals. The
detail of the map is much greater and also the areas
considered as important change in extent and loca-
tion. In this picture the important areas outside the
protected areas systems are located in correspon-
dence to the plains, to the medium altitude moun-
tains, and in Sardinia, where virtually no PA exist.
76
Brown Bear (Ursus arctos)
6.2 Mammals
77
Fig. 33a. Distribution of the irreplaceability values in Italy.
Values calculated using the distribution ranges of all Italian
mammals.
Fig. 33c. Distribution of the irreplaceability values in Italy.
Values calculated using the distribution ranges of all Italian
mammals and considering the existing Protected Areas.
Fig. 33b. Distribution of the irreplaceability values in Italy.
Values calculated using the habitat suitability models for all
Italian mammals.
Fig. 33d. Distribution of the irreplaceability values in Italy.
Values calculated using the habitat suitability models for all
Italian mammals and considering the existing Protected Areas.
Birds are with no doubt the most best group of ver-
tebrates in Italy, thus the results obtained should be
considered extremely useful. The high number of
species considered, and their great diversity of eco-
logical niches, make them one of the most interesting
groups to analyze.
Figure 34a shows the results obtained using the dis-
tribution ranges for the Italian breeding birds.
Among the important areas there are the eastern
Alps, the Apennines, and the Tyrrhenian coast, that
were considered important also for the mammals.
The analysis gives particularly great importance to
the lagoons (e.g., in Apulia, in the Padana plain,
and in Sardinia) and to some lakes. In general, the
PA system covers quite well all the areas important
for the breeding birds, but in some cases, like in
Sardinia, there is no protected area at all covering
the most important areas.
Figure 34b (built upon the breeding bird habitat suit-
ability models) depicts quite a different situation. The
spatial structure of the important areas is the same as
in Fig. 34a but their number and spatial extent
change. The importance of eastern Alps is once
more confirmed, even if the important areas are now
much more localized and fragmented. Also, in the
Apennines the number and the extent of the areas
considered important is lower than in Fig. 34a, but
the most significant change among the two maps is
that some of the lagoons are not considered impor-
tant any more, probably because they are so small
that cannot be detected in the habitat suitability mod-
els. It is important to outline also that some of the
most important areas for the Italian breeding birds
are not covered by existing protected areas, but this
gap is mostly covered by SCI and SPA.
Figure 34c shows the results obtained considering
the distribution range of the breeding birds and the
existing protected areas. It is extremely clear that the
most important areas outside the protected area sys-
tems are mostly plains, both in northern and southern
Italy. In fact, almost no Italian PA is located in flat
areas, although many of them are extremely impor-
tant for aquatic birds. Sardinia is indicated as prob-
ably the most important Italian region in this map ,
even if its territory is almost unprotected. It should be
noted that most of the areas outlined as important in
this picture are covered by SCI and SPA, which are
well located to cover to bird biodiversity.
Figure 34d shows the same general picture as Fig.
34c, but the details are different. In particular, the
Padana plain has less important areas, while central
and southern Italy, and the two largest islands (Sicily
and Sardinia) are considered of great importance. It
is important to underline that some of these areas are
not covered by any proposed protected areas, espe-
cially in northern Apulia.
Scops Owl (Otus scops)
6.3 Birds
Birds
78
6.3
6.3 Birds
79
Fig. 34a. Distribution of the irreplaceability values in Italy.
Values calculated using the distribution ranges of all Italian
breeding birds.
Fig. 34c. Distribution of the irreplaceability values in Italy.
Values calculated using the distribution ranges of all Italian
breeding birds and considering the existing Protected Areas.
Fig. 34b. Distribution of the irreplaceability values in Italy.
Values calculated using the habitat suitability models for all
Italian breeding birds.
Fig. 34d. Distribution of the irreplaceability values in Italy. Values
calculated using the habitat suitability models for all Italian breed-
ing birds and considering the existing Protected Areas.
6.4 Reptiles and Amphibians
Information on reptiles and amphibians is often limited,
especially concerning their distribution. At the same
time, many species of these two groups are endan-
gered and/or endemic of some part of the Italian
peninsula. This implies that the results obtained on
these groups should be considered with great attention
and that further investigations on their ecology and dis-
tribution should be considered before giving any con-
clusive statement.
Figure 35a depicts the results obtained from the analy-
sis of the distribution ranges. The situation outlined by
the map is unclear because of the coarseness of the
distribution maps considered. The only consideration
that can be done is that Sardinia, Liguria and the most
eastern part of Friuli are the most important areas for
Italian reptiles and amphibians.
Quite different is the situation outlined by Fig. 35b that
shows the results obtained from the habitat suitability
models for amphibians and reptiles. The differences
existing between the results obtained from the distribu-
tion ranges and those obtained from the habitat suit-
ability models are greater than those of any other
group. In fact, Fig 35b shows a much greater detail
than Fig. 35a, and also the spatial distribution of the
important areas is different. The importance of
Sardinia, Liguria and eastern Friuli is confirmed once
more, but also many areas in the Apennine chain are
classified as fundamental for amphibians and reptiles.
The results shown in Figs. 35c and 35d (from analysis
that consider the presence of protected areas) simply
confirm what has been noted above: the results
obtained from the distribution ranges (Fig. 35c) cannot
be considered really meaningful (at least for a detailed
interpretation) because of the lack of information, but
once more the importance of Sardinia, Liguria and
eastern Friuli should be underlined. On the other hand,
the results obtained from the habitat suitability models
(Fig. 35d) offer a much better vision of both the gener-
al picture and of the details, but the main conclusion is
still that Sardinia should be urgently considered in the
conservation areas of Italy.
Reptiles and Amphibians
80
6.4
Spectacled Salamander (Salamandrina terdigitata)
Western Whip Snake (Coluber viridiflavus)
6.4 Reptiles and Amphibians
81
Fig. 35a. Distribution of the irreplaceability values in Italy.
Values calculated using the distribution ranges of all Italian rep-
tiles and amphibians.
Fig. 35c. Distribution of the irreplaceability values in Italy. Values
calculated using the distribution ranges of all Italian reptiles and
amphibians and considering the existing Protected Areas.
Fig. 35b. Distribution of the irreplaceability values in Italy.
Values calculated using the habitat suitability models for all
Italian reptiles and amphibians.
Fig. 35d. Distribution of the irreplaceability values in Italy. Values
calculated using the habitat suitability models for all Italian reptiles
and amphibians and considering the existing Protected Areas.
6.5 Fish
Freshwater fish suffer the same limitations (lack of
knowledge, in this case also taxonomically) of
amphibians and reptiles, with the further limitation
that correctly representing the distribution range of a
fish is really difficult, as well as to build a habitat
suitability model for this group of vertebrates. Thus,
any conclusion that could be drawn from the analy-
sis presented in this paragraph should be considered
with great caution.
Figure 36a presents the results of the analysis car-
ried out using the distribution ranges. The map does
not allow any really useful interpretation. The only
consideration (quite obvious) is that the most impor-
tant areas for fishes in Italy are the Padana plain and
some other catchments in northern and central Italy.
Southern Italy, together with Sicily and Sardinia, has
no importance for fishes, as should be expected con-
sidering the scarcity of water bodies in those areas.
Figure 36b (results obtained using habitat suitability
models) shows quite a different view. In particular
the Adriatic region is now considered much more
important, while the detail and the spatial structure
of the important areas in the Padana plain are much
more interesting. Furthermore, the Alps, Sicily,
Sardinia and southern Italy, seem to have no impor-
tance for freshwater fishes.
It is quite interesting to note that Fig. 36c (that shows
the results obtained from protected areas and distri-
bution ranges) is extremely similar to Fig. 36a; and
Fig. 36d (results obtained from protected areas and
habitat suitability models) is similar to Fig. 36b. This
indicates that the protected areas in the Italian penin-
sula do not adeguately cover at all freshwater fish
biodiversity.
Fish
82
6.5
6.5 Fish
83
Fig. 36a. Distribution of the irreplaceability values in Italy.
Values calculated using the distribution ranges of all Italian
freshwater fish.
Fig. 36c. Distribution of the irreplaceability values in Italy. Values
calculated using the distribution ranges of all Italian freshwater fish
and considering the existing Protected Areas.
Fig. 36b. Distribution of the irreplaceability values in Italy.
Values calculated using the habitat suitability models for all
Italian freshwater fish.
Fig. 36d. Distribution of the irreplaceability values in Italy. Values
calculated using the habitat suitability models for all Italian fresh-
water fish and considering the existing Protected Areas.
The overall consideration of the analyses is that the
results are of great interest for the conservation of ver-
tebrates and Italian nature in general.
1) The method used to define the distribution models
made it possible to base the global analyses of the eco-
logical network on premises that are sound and objec-
tive. The species-specific models were constructed
using explicit, transparent methodological procedures,
which can be reiterated in the future with ever better
data sets provided by more accurate field research. It
should be underlined, once again, that one of the chief
advantages of using models such as those proposed is
that the results obtained are never, even in the worst
hypothesis, inferior to those of the classic distribution
ranges and, if all goes well, can substantially improve
our knowledge of the distribution of species.
Furthermore, model validation, mostly with positive
results, has made it possible to lay a solid, reliable
foundation for the conceptual constructions created in
drawing up the modelling process and in assessing the
system of protected areas.
2) For the first time in Italy, we now have a single data-
base, covering all species of Italian vertebrates, which
gives a few items of basic ecological information for
each, together with ranges and distribution models: the
collection, organization and standardization of the
data in digital form has resulted in a tool of great
importance, not only for ecologists but also for admin-
istrative bodies, at both central and local levels. All the
data are immediately utilizable in the IT systems most
commonly used by national, regional and provincial
administrations, and can be processed more exten-
sively and in greater detail according to the planning
needs of each institution.
3) For many species, the map of inhabited areas result-
ing from the models shows worrying fragmentation of
suitable habitats and the existence of vast unsuitable
areas, which greatly reduce or prevent the continuity of
the various components of the metapopulations. The
brief comments given for each species constitute a
starting point for more detailed analyses aimed at
obtaining results that can be directly utilized in strate-
gies and action plans targeting the species of greatest
interest for the purposes of conservation.
4) The concept of an ecological network as a dynamic
entity and as a reference for calibrating individual stud-
ies and plans for geographic areas, groups of species
or protected areas has found confirmation and support
in the analyses. The study illustrates, both graphically
and numerically, the irregular pattern of diversity for
Italian vertebrates throughout the country, identifying
the areas with the greatest richness, both for all species
taken together and for individual taxa.
5) The clear-cut characterization of Apennine Park of
Europe (APE) as a key factor for the conservation of a
large number of species represents an objective confir-
mation of the role played by the Apennines as a great
ecological corridor running through the peninsula and
of the ministerial programme aimed at giving priority
to this large area and ensuring unified intervention.
6) The fracture of environmental continuity in the band
of territory between the provinces of Matese and
Benevento has been clearly confirmed, as has the crit-
ical importance of many areas of the pre-Alps and
Apennine spurs. On the whole, the greatest diversity is
to be found in the north-eastern Alps and the northern
Apennines, but there are also areas of enormous inter-
est spread over the central and southern parts of the
country and in the great wetlands.
Stelvio National Park (Lombardy – Northern Italy)
7 Conclusions
Conclusions
7
84
7 Conclusions
7) A proper analysis of how the various species or
groups of species help to define these areas can pro-
vide information that will prove very useful in manag-
ing the protected areas and the connecting corridors,
making a valuable contribution towards the concrete
creation of the various components of the ecological
network (core areas, corridors, buffer zones), which
can be planned according to a dynamic, unified
overview of the national network but must then be cre-
ated at a local level.
8) Our analyses show the pattern of biodiversity values
for Italian vertebrates as a process of continuous, fer-
tile irradiation from the mountains to the plains; the
Alps and Apennines represent a non-metaphoric back-
bone which alone sustains much of the biodiversity of
vertebrates. From the point of view of fauna manage-
ment, this observation means two things: conservation
efforts must be concentrated on the mountainous areas
and more attention must be paid to building and man-
aging the lines of irradiation from the mountain heights
towards the plains.
9) The comparative analysis of the networks has also
led to the interesting result of showing that the network
of a few taxonomic groups or, better still, the network
of endangered species, is a good surrogate for the
global network of species. In this context, the network
of endangered species could be used as an indicator
for the global network, helping to focus attention on
top-priority species without losing sight of the overall
picture of vertebrate biodiversity.
10) The system of existing protected areas in Italy rep-
resents the various environmental categories (CORINE
Land Cover) and the various elevation belts in a very
irregular fashion, leaving important components of
Italian environmental diversity, such as hilly areas and
foothills, with little or no protection.
11) Over the last 20 years, the system of protected
areas has seen exceptional growth in the area cov-
ered; in order to conserve biodiversity, it now needs to
be consolidated within the perspective of a system inte-
grated into the surrounding territorial matrix. To this
end, it is urgently necessary to do various things: a) to
analyse the contribution made by each area to the
effectiveness of the whole system, b) to extend the
analyses to include consideration of the spatial pat-
terns of the animal populations, so as to go beyond the
simple paradigm of the presence/absence of a
species and to tackle the more complex theme of the
temporal and spatial patterns of the metapopulations,
c) to ascertain and ensure effective links between dif-
ferent areas, especially in the macro-areas that are
most critical for endangered species.
12) By identifying the high-diversity zones that are
excluded from the current system of protected areas it
is possible to plan an assessment of the entire system
and if necessary its reorganization, so as to optimise
its effectiveness, especially bearing in mind the final
stages of approval of the SCI system. In particular, the
SCI should preferably be utilized for two functions that
are critical in the context of the ecological network: a)
as corridors between the largest protected areas, or
between these and zones with lower diversity (e.g.
between mountains and plains), b) as buffer zones sur-
rounding protected areas.
13) Many important areas (as outlined by the irre-
placeability analysis) are not covered by the existing
protected areas or by the SCI and ZPS system. This
implies that the system should be revised for some
areas like the eastern Alps or part of Sardinia, in order
to consider the ecological necessities of particular
species.
14) It should be remembered that all the analysis car-
ried on during the project consider many species in
order to give a general picture of the Italian situation.
However some species (e.g. endangered wildfowl)
85
Abruzzi Chamois (Rupicapra pyrenaica ornata)
7 Conclusions
require a particular protection and should be consid-
ered singularly.
The above analyses therefore seem to give the
National Ecological Network a strong foundation, with
regard to both method and content, as a tool for plan-
ning and implementing a combination of initiatives in
the political, economic, social and territorial spheres,
aimed at a type of development that is compatible with
biodiversity conservation; this scientific basis consti-
tutes an excellent framework for renewed efforts to
develop the protected areas policy into a territorial
strategy that also includes the surrounding matrix. We
hope that the present study can represent a useful con-
tribution towards its creation.
86
Pollino National Park (Calabria – Southern Italy)
8 Bibliography
Amori G., Angelici F. M. & L. Boitani. 1999.
Mammals of Italy: a revised checklist of species and
subspecies. Mammalis 79 (2): 271-286.
Andreotti A., Baccetti N., Perfetti A., Besa M.,
Genovesi P. & V. Guberti. 2001. Mammiferi e Uccelli
esotici in Italia: analisi del fenomeno, impatto sulla bio-
diversità e linee guida gestionali. Quad. Cons.
Natura, 2 Min. Ambiente – Ist. Fauna Selvatica.
Battisti C. 2002. Reti ecologiche e fauna: approcci
strutturali, funzionali, gestionali – un contributo
metodologico. Parchi now printing.
Bianco P. G. 1998. Freshwater fish transfers in histo-
ry, local changes in fish fauna, and a prediction on the
future of native populations. In I. G. Cowx (ed.)
Stocking and introduction of fish. Fishing New Books,
Blackwell Science Ltd., Oxford.
Boitani L. 2000. Rete ecologica nazionale e conser-
vazione della biodiversità. Parchi 29: 66-74.
Boitani L., Corsi F., Falcucci A., Marzetti I., Masi M.,
Montemaggiori A., Ottavini D., Reggiani G. & C.
Rondinini. 2002. Rete Ecologica Nazionale. Un
approccio alla Conservazione dei Vertebrati Italiani.
Relazione Finale; Ministero dell’Ambiente e del
Territorio.
Brichetti P. & A. Gariboldi (Eds.). 1997. Manuale
pratico di ornitologia. Ed. Agricole, Bologna.
Brichetti P. & B. Massa. 1998. Check list degli uccel-
li italiani aggiornata a tutto il 1997. Riv. Ital. Orn. 68
(2): 129-152.
Bulgarini F., Calvario E., Fraticelli F., Petretti F. & S.
Sarrocco. (eds), 1998. Libro Rosso degli Animali
d’Italia – Vertebrati. WWF Italia, Roma.
Dobson A., Rodriguez J. P., Roberts W. M. & D. S.
Wilcove. 1997. Geographic distribution of endan-
gered species in the United states. Science 275: 550-
553.
Dunning J. B., Danielson B. & H. R. Pulliam. 1992.
Ecological processes that affect population in complex
landscapes. Oikos 65 (1): 169-175.
Duprè E. 1996. Distribuzione potenziale del lupo
(Canis lupus) in Italia e modelli di espansione del-
l’rangee: un approccio multivariato sviluppato attra-
verso un GIS. Tesi di Dottorato in Biologia animale.
Università degli Studi di Roma “La Sapienza”.
Ferrier S., Pressey R.L. & T.W. Barrett. 2000. A new
predictor of the irreplaceability of areas for achieving
a conservation goal, its application to real-world plan-
ning, and a research agenda for further refinements.
Biological Conservation 93: 303 – 325.
Gandolfi G., Zerunian S., Torricelli, P. & A.
Marconato. 1991. I pesci delle acque interne italiane.
Istituto Poligrafico e Zecca dello Stato, Roma.
Gambino R. & G. Negrini. 2001. Il sistema
nazionale delle aree protette nel quadro europeo: clas-
sificazione, pianificazione e gestione. Rapporto Finale
al Ministero dell’Ambiente e della Tutela del Territorio.
Gustafson E. J. & R. H. Gardener. 1996. The effect
of land heterogeneity on the probability of patch colo-
nization. Ecology 77: 94-107.
Kerr J. 1997. Richness, endemism and the choice of
areas for conservation. Conservation Biology 11 (5):
1094-1100.
Kottelat M. 1997. European Freshwater fishes.
Biologia 52 (suppl.): 1-152.
Bibliography
8
87
8 Bibliography
Manson R. H., Ostfeld R. S. & C. D. Canham. 1999.
Responses of small mammal community to heterogene-
ity along forest old field edge. Landscape ecology 14:
355-367.
IUCN. 2001. The Red List of Endengered Species.
www.redlist.org.
McGarigal K. & B. J. Marks. 1999. FRAGSTATS
Spatial Pattern Analysis Program for Categorical
Maps. University of Massachusetts, Amherst, MASS.
http://www.umass.edu/landeco/research/fragstats/f
ragstats.html.
New South Wales National Parks and Wildlife
Service. 2001. C-PLAN. Conservation planning soft-
ware. User manual for C-Plan Version 3.06.
http://www.ozemail.com.au/~cplan.
Pressey R.L., Johnson I.R. & P.D. Wilson. 1994.
Shades of irreplaceability: towards a measure of the
contribution of sites to a reservation goal. Biodiversity
and Conservation 3: 242 – 262.
Pressey R.L., Ferrier S., Hutchinson C.D., Sivertsen
D.P. & G. Manion. 1995. Planning for negotiation:
using an interactive geographic information system to
explore alternative protected areas networks. Pp 23 –
33 in Saunders D.A., Craig J.L. & E.M. Mattiske (Eds.).
Nature conservation: the role of networks. Surrey
Beatty and Sons, Sidney.
Oliver I., Beattie A. J. & A. York. 1998. Spatial fideli-
ty of plant, vertebrates and invertebrates assemblages
in multiple use in forest in eastern Australia.
Conservation Biology 12 (4): 822-835.
Reggiani G., Boitani L. & G. Amori. 2000. I con-
tenuti ecologici di una rete ecologica. Atti del conveg-
no “Reti ecologiche azioni locali di gestione territoriale
per la conservazione dell’ambiente”. Centro Studi
Valerio Giacomini, Gargnano 12-13 October 2000.
Societas Herpetologica Italica, 1996. Atlante provvi-
sorio degli Anfibi e dei Rettili italiani. Ann. Mus. Civ.
di St. Nat. G. Doria, vol. XCI: 95-178, 1996.
Tucker G.M. & M.F. Heath (Eds.). 1994. Birds in
Europe. Their conservation status. BirdLife
International, Cambridge.
Williams P., Gibbons D., Margules C., Rebelo A.,
Humphries C. & R. Pressey. 1996. Comparison of rich-
ness hotspots, rarity hotspots and complementary
areas for conserving biodiversity, using British birds.
Biological Conservation 10 (1): 155-174.
88
IEA
Institute of Applied Ecology
Via L. Spallanzani, 32 - 00161 Rome - Italy
Tel./fax: +39 06 4403315 - e-mail: iea@ieaitaly.org
In collaboration with:
ISBN 88-87736-03-0
... Therefore, it appears as a continuum of protected areas, which would make any resulting ecological network, made up of short corridors, completely useless. In order to evaluate the different structure of the two ecological networks , the network comparison analysis was carried out by referring to tree synthetic indexes calculated for both study areas: i) landscape fragmentation (LF): mainly referred to the habitat fragmentation and isolation phenomena (in structural as well as in functional terms) and caused by the interference of anthropic activity (Jaeger, 2000; Romano, 2003 Romano, , 2005); ii) environmental sustainability index (ESI): an indicator of ecological stability defined starting from the balance of pressure effects and the regeneration potential typical of each type of land use (Magoni and Steiner, 2001); iii) habitat suitability index (HSI): an aggregate measure of the territorial quality level of the species-environment relationships, defined for the Italian vertebrate by Boitani et al. (2003). The indices analysis (except the HSI) and the implementation of the two ecological networks were based on 1:100,000 Corine land cover as reference map. ...
... What highlighted does not mean that Natura 2000 sites should be excluded from the ecological network. Indeed, their role on biodiversity conservation is widely recognised and remarked by scholars (Boitani et al., 2003; Biondi et al., 2012; Modica et al., 2012). Rather, this issue opens new perspectives of reflection and therefore new challenges in their sustainable planning and management. ...
... In 2002, the Italian Ministry of the Environment (Nature Conservation Service) promoted a study aimed at the identification and analysis of the conservation needs of the Italian vertebrates (Boitani et al., 2003). Still today, this is the only research carried out at national scale in Italy, and provides information on biodiversity and on habitat quality for most Italians native terrestrial vertebrates (33 amphibians, 92 mammals, 207 breeding birds, 37 reptiles, and 45 freshwater fishes were scrutinised). ...
Article
Full-text available
The issue of the fragmentation of natural habitats is increasingly at the core of the scientific debate, but often not taken into due consideration in planning tools, with particular reference to the dynamism and complexity of landscapes. As it has been recognised at a European level, in order to allow different species to remain in good functional status, a network of green infrastructures should be implemented. The concept of ecological island is no longer sufficient to protect adequately the fauna and the ecosystem it lives in. Consequently, ecological islands must turn into ecological networks. The ecological connectivity refers to the way habitats are physically connected to each other and how easy it is for species to move in. Good ecological connectivity is fundamental to the effective conservation of biodiversity considering that most species and ecological functions provided by ecosystems (ecosystem services) require a much wider space than that available within the boundaries of a single protected area. The main objective of this paper is to critically compare the application of a dedicated model for the design of ecological networks to two very different environmental contexts. This model was first tested in a Mediterranean area (the province of Reggio Calabria) in 2008 and updated in 2010. The main goal was to integrate the traditional (physiographic and functional) approaches into the design of ecological networks by taking into account biological and orographic elements as well as the anthropic structure of the territory. In 2011, within the ECONNECT European project, the same model was applied to the Pilot region of South- Western Alps (including the French region of Provence-Alpes-Côte d’Azur and the Italian regions Piedmont and Liguria), which is one of the richest transnational districts in Europe in terms of biodiversity. In such a region, the issue of multidisciplinary ecological connectivity was tackled in order to provide a series of proposals aiming at the development of the ecological potential of the area. The two applications allowed to further investigate the strengths and weaknesses of the implemented model by integrating its validation also by means of information on faunal presence, which obviated one of the major limitations occurred in the first application.
... Il sistema che ne risulta complessivamente presenta alcuni elementi di disequilibrio nella rappresentatività dei sistemi di paesaggio, degli ecosistemi presenti nella Regione e quindi, presumibilmente, nella rappresentazione della biodiversità (Tallone e Arcà, 2002), che sono stati anche fortemente criticati dal punto di vista della efficacia per la conservazione (Battisti e Gippoliti, 2004). @BULLET Valutazione del valore delle unità ambientali regionali per l'ulteriore aggiunta di rappresentatività del sistema, attraverso l'uso di software - " Irreplaceability analisys " ; @BULLET Valutazione, per alcune specie chiave, dell'efficacia a lungo termine del sistema di aree protette per evitare il rischio di estinzione anche attraverso l'analisi di modelli predittivi come le PVA (Population Viability Analysis); @BULLET Collegamento delle suddette analisi regionali a quelli della Rete Ecologica Nazionale effettuata a cura del Ministero dell'Ambiente (Boitani et al., 2003); @BULLET Informazione e divulgazione nell'ambito del sistema delle aree protette (personale ed amministratori ) dei contenuti scientifici dei temi suddet- ti. 2. Analisi delle questioni relative alla classificazione delle aree protette (Thomasset, Castelnovi e Gambino , in prep.), in riferimento allo sviluppo di tali temi alla scala internazionale, con una applicazione alla scala regionale degli studi già effettuati in precedenza per il Ministero dell'Ambiente (Gambino, 2002); 3. Integrazione della cartografia dell'uso del suolo disponibile per la Regione Lazio, con riferimento ad aree campione di specifico interesse per la pianificazione sistematica delle aree protette, ed in particolare ad una fascia sul gradiente mare-montagne tra il litorale nord della Provincia di Roma, la Tolfa, il Tevere e i Monti Reatini; 4. Approfondimento degli studi territoriali e sulla biodiversità per alcune aree campione della Regione Lazio potenzialmente di particolare interesse per la istituzione di nuove aree protette: Monti Lepini e fascia Farfa-Tevere-Treja (Agliata, Cingolani e Leoni, in prep). L'insieme di tali approfondimenti potrà dare un quadro più dettagliato delle necessità di revisione del sistema delle aree protette al fine di una migliore rappresentatività della biodiversità. ...
... L'insieme di tali approfondimenti potrà dare un quadro più dettagliato delle necessità di revisione del sistema delle aree protette al fine di una migliore rappresentatività della biodiversità. Boitani et al. 2002 Banca dati CK Map (Ministero Ambiente) Boitani et al. 2003 INFS (1995Attività sistematica INFS e Progetto Piccole Isole (Spina, in verbis) Atlante Nazionale degli Anfibi e Rettili ...
... I patterns di rappresentazione basati su semplici osservazioni di presenza (nuvole di punti) sono relativamente efficienti nel descrivere le esigenze di specie sedentarie, mentre per specie più mobili la questione diventa più complessa. Per l'intero territorio italiano Boitani et al. (2003) , cercando di superare i problemi relativi alla descrizione cartografica della presenza della fauna, per rendere più dettagliati i semplici modelli descrittivi tramite areale di distribuzione hanno definito modelli di idoneità ambientale per tutti i vertebrati italiani, validati con osservazioni puntuali . Tali modelli aiutano notevolmente nel definire la capacità delle diverse aree protette di rappresentare la presenza delle specie di vertebrati, ed anche i patterns di insostituibilità. ...
Conference Paper
Full-text available
Conservation planning of Regione Lazio protected areas in the local legal framework.
... Therefore, it appears as a continuum of protected areas, which would make any resulting ecological network, made up of short corridors, completely useless. In order to evaluate the different structure of the two ecological networks , the network comparison analysis was carried out by referring to tree synthetic indexes calculated for both study areas: i) landscape fragmentation (LF): mainly referred to the habitat fragmentation and isolation phenomena (in structural as well as in functional terms) and caused by the interference of anthropic activity (Jaeger, 2000; Romano, 2003 Romano, , 2005); ii) environmental sustainability index (ESI): an indicator of ecological stability defined starting from the balance of pressure effects and the regeneration potential typical of each type of land use (Magoni and Steiner, 2001); iii) habitat suitability index (HSI): an aggregate measure of the territorial quality level of the species-environment relationships, defined for the Italian vertebrate by Boitani et al. (2003). The indices analysis (except the HSI) and the implementation of the two ecological networks were based on 1:100,000 Corine land cover as reference map. ...
... What highlighted does not mean that Natura 2000 sites should be excluded from the ecological network. Indeed, their role on biodiversity conservation is widely recognised and remarked by scholars (Boitani et al., 2003; Biondi et al., 2012; Modica et al., 2012). Rather, this issue opens new perspectives of reflection and therefore new challenges in their sustainable planning and management. ...
... In 2002, the Italian Ministry of the Environment (Nature Conservation Service) promoted a study aimed at the identification and analysis of the conservation needs of the Italian vertebrates (Boitani et al., 2003). Still today, this is the only research carried out at national scale in Italy, and provides information on biodiversity and on habitat quality for most Italians native terrestrial vertebrates (33 amphibians, 92 mammals, 207 breeding birds, 37 reptiles, and 45 freshwater fishes were scrutinised). ...
Article
The issue of the fragmentation of natural habitats is increasingly at the core of the scientific debate, yet it is not taken into account in planning tools, with particular reference to the dynamism and complexity of landscapes. As it has been recognised at a European level, in order to enable different species to remain in good functional status, a network of green infrastructures is required. The concept of “ecological island” is no longer sufficient to adequately protect the fauna and the ecosystem it lives in. As a consequence, ecological islands must turn into ecological networks. The Ecological connectivity refers to the way habitats are physically connected to each other and how easy it is for species to move in. Good ecological connectivity is fundamental to the effective conservation of biodiversity considering that most species and ecological functions provided by ecosystems (ecosystem services) require a much wider space than that available within the boundaries of a single protected area. The main objective of this paper is to critically compare the application of a model for the design of ecological networks to two very different environmental contexts. This model was first tested in a Mediterranean area (the Province of Reggio Calabria) in 2008; the goal was to integrate the traditional (physiographic and functional) approaches into the design of ecological networks by taking into account biological and orographic elements as well as the anthropic structure of the territory. In 2011, within the ECONNECT European project, the model was applied to the pilot region of South-Western Alps (including the French region of Provence-Alpes - Côte d’Azur and the Italian regions Piedmont and Liguria), which is one of the richest transnational districts in Europe in terms of biodiversity. In such a region, the issue of multidisciplinary ecological connectivity was tackled in order to provide a series of proposals aiming at the development of the ecological potential of the area. The two applications allowed to further investigate the strengths and weaknesses of the implemented model by integrating its validation with information on faunal presence, which obviated one of the major limitations occurred in the first application.
... Because "prey availability" may represent a variable associated with road-kills (Barrientos & Bolonio, 2009;Barrientos & Miranda, 2012), we considered the habitat suitability models provided by the National Ecological Network (NEN) project (Boitani, Falcucci, Maiorano, & Montemaggiori, 2003) for 26 potential polecat prey species (Baghli, Walzberg, & Verhagen, 2005;Lodé, 1997;Prigioni & De Marinis, 1995;Sidorovich & Pikulik, 1997; see Appendix S1). The NEN rasters are provided at a resolution of 100 × 100 m grid cells, ...
... Oryctolagus cuniculus) is by far the main prey in Spain (Barrientos & Bolonio, 2009), which makes it easier to model the influence of prey abundance on the probability of road-kill. Alternatively, the field sampling carried out by Barrientos and Bolonio (2009) could be more accurate than the map of habitat suitability for prey species (Boitani et al., 2003) that we used in the present study. ...
Article
Full-text available
Aim Roads impact wildlife in different ways, among which road mortality has been the most studied. Budgets in conservation biology are usually small, and macroecological approaches have been employed in recent years as the first steps towards guiding management. Carnivores are particularly vulnerable to mortality on roads due to their elevated ecological needs (low population density, often low fecundity and relatively large home ranges). Our aim was to develop a ranking methodology to prioritize specific areas for road‐kill mitigation. Location Continental Italy. Methods We studied 271 occurrences of live polecats (Mustela putorius ) and 212 polecat road‐kill sites. We used the former to generate a species distribution model and the latter to identify the variables that determined the road‐kill risk. Habitat suitability was derived from a spatial distribution model that combined the polecat occurrence data with a set of environmental variables. Prey availability was derived from the combination of suitability maps of 26 prey species. We used generalized linear modelling to identify the set of variables that best explained the occurrence of road‐kills. The variables included in the best performing model were combined to produce the road risk map and to identify the areas with the highest densities of road sections with highest risk. Results Road‐kills were positively associated with the road sections with higher broad‐leaved forest coverage. The number of casualties was found to be higher than expected on the national and provincial roads and lower than expected on the local roads. Main conclusions This approach allowed us to identify the 10 × 10 km cells where mitigation actions to prevent road‐kills should be prioritized. As mitigation actions (wildlife passage construction, fencing) are expensive, measures should be prioritized on the specific high‐risk road sections inside these selected cells, avoiding generalized mitigation plans.
... The fact that habitat fragmentation represents one of the main threats to biodiversity at a continental, regional and landscape level (Davies et al., 2001;Haila, 2002;Fahrig, 1997;Fahrig, 2003;Lindenmayer and Fisher, 2006) has led to the conclusion that connectivity, as a concept, and ecological networks, as an operational tool, are a major answer to maintain healthy ecosystems and biodiversity (Linehan et al., 1995;Rosenberg et al., 1997;Bennett, 1998;Battisti, 2003;Boitani et al., 2003). ...
... Some mountains regions of Italy (i.e. the Apennines) show a natural tendency to serve as large-scale biological areas for connectivity. For example, although the Abruzzi Region is well known for the large protected areas suitable as source for large mammals, none of these reserves except the Abruzzo Lazio and Molise National Park serve as a stable refuge for a small breeding population of Apennine brown bear (Boitani et al., 2003;Posillico et al., 2004;Ciucci and Boitani, 2008;Amori et al., 2009), as only male bears are wandering among Apennine protected areas. It seems, therefore, that the true problem for this species and many others may be the quality and size of the source populations, in core areas and the biology and social system of the species; not the lack of connectivity per se. ...
Article
The paper critically examines the equivoques, conceptual traps and weaknesses of the recent 'ecological network' paradigm, invocated as a cool environmental planning tool to the aim to mitigate the effect of habitat fragmentation on biodiversity. We highlight as: (i) there is a semantic ambiguity deriving from the languages used in this interdisciplinary arena; (ii) these plans will be considered a true tool for biodiversity conservation when they will adopt a logic of problem solving and the standards requested in project cycle management (clear objectives, decision-making approach, appropriate monitoring and indicators, adaptive management); (iii) planners should follow a costs/benefits analysis comparing different scenarios and verifying that the 'connectivity' option effectively work better; (iv) each ecological network should be considered as a context-specific strategy where connectivity is only a simplified and schematic key of interpretation; (v) planners should carried out a local selection of fragmentation-sensitive targets that may not correspond with the species of conservation concern included in global or national red lists.
... In 1999, the Nature Conservation Service of the Ministry of the Environment published an Interim Report of the Sectoral Working Group "National Ecological Network" offering a definition of the concept of ecological network and identifying a number of useful measures for the emergence of such a structure at the national level (Ministero dell'Ambiente, 1999). Following this report, a national ecological network was mapped (Boitani et al., 2003). In addition, the Italian National Institute for Environmental Protection and Research (ISPRA) has monitored the inclusion of ecological connectivity issues in regional legislative frameworks and spatial planning processes at regional and local levels (Guccione and Schilleci, 2010). ...
Article
Against the background of widespread biodiversity loss, the issue of ecological connectivity has become increasingly important in recent decades at EU level. This growing attention can be observed in particular in biodiversity strategies or certain directives, in the establishment of key programmes or in the mobilisation of various types of funds. However, there is currently no EU framework that explicitly obliges member states to address the issue or to achieve certain objectives in this field. This article first shows how ecological connectivity has gradually emerged as a spatial planning issue, in particular through national/federal and/or regional legislative developments, in different countries (Germany, Italy and France). A focus on the French context and its Green and Blue Network Policy then depicts the implementation challenges observed in spatial planning at the regional, sub-regional and local levels. The paper highlights the difficulty of establishing a standardised framework for ecological connectivity, which implies considering the complexity of wildlife in spatial planning procedures in which potentially divergent interests intersect.
... b) Anfibi e rettili -Dati dell'Atlante "Anfibi e rettili del Lazio" (Bologna et al., 2000) e localizzazione di siti ivi segnalati. c) Mammiferi -Modelli di distribuzione della REN di Boitani et al. (2002Boitani et al. ( , 2003. d) Invertebrati -Dati preliminari dell'Osservatorio sulla Biodiversità (ARP/Università Tor Vergata/altre Università regionali) in particolare su invertebrati (Lepidotteri, Coleotteri Carabidi) e animali troglobi. ...
Chapter
Full-text available
Legal and technical framework for protected area conservation planning in Lazio Region, Italy.
Thesis
p>The present study is an effort towards the international and multidisciplinary approach to conservation of European biodiversity. The main aim was to map the distribution of suitable areas for the conservation of bears, lynx and wolves in the Carpathian Mountains. It was done applying a distance classifier, the Mahalanobis distance, over a set of environmental variables representing the region. The results suggested that 41, 58 and 65% of the Carpathian Ecoregion is highly suitable for bear, lynx and wolf, respectively. Considering the three carnivores at once, 20% of the area is highly suitable. Suitable areas are fragmented, but interspersed with areas of less suitability value, without being isolated, and spatially distributed all along the Mountain range. The results were validated with an independent data set and results suggest that the model produced an acceptable estimate of the areas effectively occupied by the carnivores. The comparison between suitability maps obtained with the two independent data sets showed that they were consistent, always reaching values of K-statistics > 0.5. The development of human activities over the land poses problems of how to integrate land exploitation and biodiversity conservation. The outputs of the environmental modelling exercise were used for estimating the distribution of potential conflicts between the presence of carnivores and livestock husbandry practices. Results suggested an effective management would avoid the summer grazing of livestock in carnivore areas and the use of damage prevention measures. The actual effect of currently protected areas in the region was assessed and the need of an increased portion of protected land, particularly in Romania and Ukraine emerged after analysing the proportion of highly suitable areas for large carnivores under any kind of legal protection.</p
Article
Full-text available
The effect of landscape heterogeneity on the dispersal of organisms between habitat islands is poorly understood. Preferred pathways for dispersal (i.e., corridors), as well as dispersal barriers, are difficult to identify when the landscape matrix is composed of a complex mixture of land cover types. We developed an individual-based dispersal model to measure immigration and emigration rates between habitat islands within heterogeneous landscapes. Dispersing individuals of a model organism were simulated as self-avoiding random walkers (SAW) traversing a digital land cover map, with each habitat type assigned a priori a probability that the SAW would enter that habitat type. Each individual began the dispersal process on a random site at the edge of a deciduous forest patch and was allowed to move until it reached a different deciduous forest patch. Visualization of the movement patterns across the landscape was achieved by tabulating the frequency of visitation of successful dispersers to each grid cell on the map. The model was used to estimate the probabilities of disperser transfer between patches by varying the a priori probabilities of movement into each habitat type in order to: (1) estimate the effect of changing landscape heterogeneity on the transfer probabilities, and (2) visualize dispersal corridors and barriers as perceived by model organisms operating by specific movement rules and at specific scales. The results show that 89% of the variability in dispersal success can be accounted for by differences in the size and isolation of forest patches, with closer and larger patches having significantly greater exchange of dispersing organisms. However, changes in the heterogeneity of the landscape matrix could significantly enhance or decrease emigration success from an individual patch, depending on the landscape. Changes in emigration success from an individual patch resulting from changes in matrix heterogeneity were not predictable, and transfer rates between patches were not symmetrical due to differences in the proximity of neighboring patches, and differences in the funneling attributes of certain landscape patterns. Visualizations showed that corridors are often diffuse and difficult to identify from structural features of the landscape. A wide range of organisms with differing movement capabilities can be simulated using the approach presented to increase our understanding of how landscape structure affects organism dispersal.
Article
Full-text available
Despite the importance of edges effects in ecological systems, the causes and consequences of animal responses to habitat edges are largely unknown. We used three years of live-trapping and measures of the plant community around trap stations to explore the responses of white-footed mice (Peromyscus leucopus), meadow voles (Microtus pennsylvanicus), and short-tailed shrews (Blarina brevicauda) to forest-field edges in upstate New York. We found that capture probabilities of voles were highest in grass- and forb-dominated micro-habitats and in old-field zones distant from the forest edge. In contrast, capture probabilities of white-footed mice were highest in shrub-dominated microhabitats and in zones near the forest edge. Short-tailed shrews did not show strong micro- or macrohabitat associations. The responses by voles, the competitive dominant in our system, to variation along forest-field edges were more consistent across years than were those of the competitively inferior, white-footed mouse. Mice were less likely to use the old-field interiors when vole density was high than when it was low, suggesting competitive displacement of mice by voles. Finally, we found good agreement between the spatial activity patterns of mice and voles in old-fields and their impacts on patterns of survival of tree seeds and seedlings in concurrent studies. These results suggest that a dynamic interaction exists between the plant and animal communities along forest edges.