The status of Sarcopoterium spinosum (Rosaceae) at the western

Israel Journal of Plant Sciences
Vol. 55
2007
pp. 1–13
This paper has been contributed in honor of Doctor No’am Seligman.
The status of Sarcopoterium spinosum (Rosaceae) at the western periphery of its range:
Ecological constraints lead to conservation concerns
Domenico Gargano,a,* Giuseppe Fenu,b Piero Medagli,c Saverio Sciandrello,d and Liliana Bernardoa
Museo di Storia Naturale della Calabria ed Orto Botanico dell’Università della Calabria, Loc. Polifunzionale,
I-87030 Arcavacata di Rende (CS), Italy
b
Centro Conservazione Biodiversità (CCB), Dipartimento di Scienze Botaniche dell’Università degli Studi di Cagliari, viale S. Ignazio da Laconi 13, I-09123 Cagliari, Italy
c
Dipartimento di Scienze e Tecnologie Biologiche ed Ambientali dell’Università del Salento, via Monteroni, I-73100 Lecce, Italy
d
Dipartimento di Botanica dell’Università degli Studi di Catania, via A. Longo 19, I-95125 Catania, Italy
a
(Received 6 July 2007 and in revised form 16 December 2007)
Abstract
This paper considers the demography, ecology, and conservation perspectives of the
western populations of the clonal dwarf-shrub Sarcopoterium spinosum (L.) Spach.
While in the Middle East the species dominates large areas, westwards it is less
common and many populations have become extinct since the late 19th century. We
highlight the ecological limitations and their implications for the conservation of the
species at the periphery of its range. In southern Italy we studied the demographic
traits of two populations and the age spatial structure of a third. Ramet density and
lifespan appear to be lower than in the Middle East. The floristic analysis of the
stands studied and a climatic analysis over the whole range of S. spinosum provide
a key for the interpretation of such differences. Given that summer drought stress
decreases westwards, both the sprouting vigor and the ecological space available for
S. spinosum become limited by increasing competition. This makes the populations
more likely to become extinct in changing landscapes, as revealed by the decreasing
extent of occurrence and area of occupancy due to habitat loss. Although S. spinosum is not at risk in the Middle East, at the western border of its range it qualifies as
an endangered species.
Keywords: clonal, conservation, demography, habitat changes, peripheral populations, Sarcopoterium spinosum
Introduction
Border populations represent an interesting target for
ecological, evolutionary, and conservation issues (Grant
and Antonovics, 1978; Lesica and Allendorf, 1995; Holt
and Keitt, 2005). Many authors (Grant and Antonovics,
1978; Volis et al., 1998; Nantel and Gagnon, 1999; Lönn
and Prentice, 2002; Médail et al., 2002; Shea and Furnier, 2002; Busch, 2005) have demonstrated that plant
traits (e.g., ecology, morphology, demography, breeding
system, and genetic settings) vary across the species’
range, which is termed the central/marginal concept.
Such variation is the main factor responsible for the relevancy of the border populations to micro-evolutionary
processes (Grant and Antonovics, 1978; Moritz, 2002).
Indeed, Thompson (2005: 107) writes “Somewhere at
the limits to the distribution of widespread species, new
*Author to whom correspondence should be addressed.
E-mail: [email protected]
© 2007 Science From Israel / LPPltd., Jerusalem
taxa are evolving that will become the rare endemics
of future conservation efforts.” It is important to note
that the population traits do not vary in a univocal
fashion between peripheral and core areas. In fact, the
local ecological conditions have a great influence on
population processes at the range’s margins, by conditioning the patterns of selection and plant responses
(Jonas and Geber, 1999; Volis et al., 2000; Herlihy and
Eckert, 2005). As a consequence, general assumptions
(e.g., concerning reduced species abundance and fitness
toward the range’s edges) may be wrong (Sagarin and
Gaines, 2002, 2006). On the other hand, reduced genetic
diversity and higher extinction rates have often been
found in border populations (Lammi et al., 1999; Nantel
and Gagnon, 1999; Lönn and Prentice, 2002; Lienert et
al., 2002; Shea and Furnier, 2002). This appears to be
due mainly to the less favorable ecological conditions
that plant populations are likely to face at the margins of
their range (Thomas et al., 2001; Holt and Keitt, 2005).
Hence, suitable habitats become less frequent and the
individual patches are smaller than in the core areas,
reducing the connectivity among populations (Channell
and Lomolino, 2000; Ezard and Travis, 2006). As a consequence, increasing isolation and smaller population
size make the populations more prone to genetic drift
and inbreeding depression (Lönn and Prentice, 2002).
In this paper we analyze ecological constraints related to population trends and the conservation status
of Sarcopoterium spinosum (L.) Spach at the western
periphery of its range. This clonal dwarf-shrub probably
originated at the eastern margins of the Mediterranean
basin, and then expanded westwards (Litav and Orshan,
1971; Zohary, 1973). Currently, according to Proctor
(1968) and Martinoli (1969), in the W Mediterranean
basin the species occurs in locations on the Balkan
Coast, in Italy, and in Tunisia (Djerba). In addition,
a small population of S. spinosum is also recognizable in Malta (S. Brullo, unpublished data), where the
plant suffers from reproductive limitation (J. Buhagiar,
pers. comm.). Based on the summarized data, the species reaches the western limit of distribution in Italy,
with the populations in Sardinia. From an ecological
viewpoint, S. spinosum communities dominate a wide
spectrum of habitats in the Middle East; they include
the batha vegetation of the Mediterranean coast as well
as the semisteppic plant aggregates of the hinterlands
(Zohary, 1973; Chiesura Lorenzoni and Lorenzoni,
1977). In these areas the species shows a great ability
to persist and expand even at high disturbance (Litav
and Orshan, 1971; Seligman and Henkin, 2000, 2002).
Along the Eastern Mediterranean coast, as described
for Greece (Lavrentiades, 1969; Chiesura Lorenzoni
and Lorenzoni, 1977), the main vegetation type hosting
Israel Journal of Plant Sciences 55
2007
S. spinosum is the phrygana. In Italy the plant occurs
in few regions (Pignatti, 1982), inhabiting a narrow
range of habitats in isolated and generally small areas
close to the shoreline. Therefore, it is mainly found in
garigues communities with a phrygana-like structure
(Biondi and Mossa, 1992). Such communities are linked
to the residual coastal vegetation described as Genisto
corsicae–Sarcopoterietum spinosi by Biondi and Mossa
(1992) for Sardinia, Cisto monspeliensis–Sarcopoterietum spinosi by Brullo et al. (1997) for Apulia, and
Helichryso italici–Sarcopoterietum spinosi by Géhu et
al. (1984) for Calabria. The latter includes the subassociation nerietosum oleandri, typical of large stream
beds in northeastern Calabria (Biondi et al., 1994). Only
in SE Sicily does the species appear able to colonize
hinterlands, within phrygana-like communities whose
expansion is favored by grazing and fire (Barbagallo et
al., 1979; Bartolo et al., 1985). Many local extinctions
of S. spinosum have been recorded in the Italian regions
(except for Sicily) where it was once present (Caniglia
et al., 1974; Pignatti, 1982), and similar events are
recognized in other western areas (Balkan coast) of the
species range (Martinoli, 1969).
A variation in species trends between eastern and
western areas of its range appears evident. It was hypothesized that ecological limitations resulting from the
geographical position of the populations may be responsible for such differences and thus have implications for
the conservation of S. spinosum at the western periphery
of its range. To corroborate our hypothesis, we analyzed
the population age structure, flora, and climate in some
S. spinosum stands in Italy. By comparing with information on the Middle East areas, we found evidence
for (a) reduced sprouting vigor and competition ability,
mainly due to climatic constraints, and (b) the role of
such limitations, together with anthropogenic or natural
habitat changes, in threatening the species persistence
at a regional scale.
Materials and Methods
Study sites
As mentioned above, in Italy the main habitat types of
S. spinosum seem to be garigues/scrublands occurring
near the coast (the most frequent condition) or towards
the hinterland (limited to Sicily). We chose to study
population establishment under different ecological
(coast/hinterland) and management scenarios (different
land use/protection regimes). Therefore S. spinosum
was sampled in three stands (Fig. 1), located in SE Sicily, NE Calabria, and SW Apulia (hereafter referred to as
SI, CA, and AP). The SI stand (Nebrodi Mountains near
Buccheri; 37°09¢N, 15°02¢E) is located at 560 m asl, in
a hilly area with calcareous bedrock; the site is disturbed
by fire and grazing; the vegetation (Chamaeropo humiliSarcopoterietum spinosi Barbagallo, Brullo and Fagotto,
1978) is dominated by cushions of nanophanerophytes
and chamephytes. In contrast, the CA stand (Fiumara
Canna; 40°07¢N, 16°35¢E) is in an undisturbed site at
less than 30 m asl in a large stream-bed (fiumara) near
the Ionian Sea; S. spinosum is included in plant communities established on rocky sediments (Helichryso
italici-Sarcopoterietum spinosi nerietosum oleandri).
Finally, the AP stand (Palude del Capitano; 40°17¢N,
17°55¢E) is located in a natural reserve at 10 m asl, on a
limestone bedrock; although the mediterranean maquis
dominates the stand, where the shrubs are less dense, S.
spinosum occurs in communities belonging to the Cisto
monspeliensis–Sarcopoterietum spinosi.
Demography
Our work was carried out according to the protocol of
Seligman and Henkin (2002) for collecting and interpreting demographic data. For SI and CA we measured
Fig. 1. Italian regions (in gray) with current occurrence of S. spinosum, sampled populations (1—Sicily; 2—Calabria; 3—Apulia),
and land-use maps for habitat trend analysis (arrow indicates position of mapped area on northeastern coast of Calabria).
Gargano et al. / Conservation ecology of S. spinosum W-populations
the height and the canopy size area (using the average
diameter of two diameters measured at right angles) of
the discrete cushions located in four square plots (1 ´
1 m) placed at random. We then uprooted the measured
cushions. Since the population of AP (the only one living in Apulia) was found to be very poor, we avoided uprooting. However, in this stand we measured the height
and canopy size of all the discrete cushions (49) along
a transect of 220 ´ 20 m from the coastline towards the
hinterland (the start and end points of the transect were
located near the two cushions of S. spinosum that were
farthest apart).
In the laboratory the uprooted material was examined
by identifying the individual plants, the ramets already
independent (RAMs), and the clonal units connected to
the mother plant (SRAMs); the latter element includes
rooting stems with connection to other ramets. The
cross section of the taproot was measured for 350 units
and then, in the transition zone between root and stem,
we obtained smooth sections to count the growth rings.
In accordance with Seligman and Henkin (2002), the
elements without visible rings were considered to be
dead. The age of the whole cushion was assumed to correspond to the oldest RAM/SRAM. Regression analysis
was performed to evaluate the relationship between age
and canopy size of the cushions. The fitted regression
model for cushion age vs. canopy size was used to estimate the age of the cushions in AP. By calculating the
level of spatial autocorrelation for the age variable, such
estimations were used to check for a spatial structure
along the transect. We performed further investigations
on the variation pattern of the cushions’ age in AP by
means of regression analysis. In order to estimate the
age of dead RAMs and SRAMs, we fitted a regression
model to test the relationship between age and taproot
cross section diameter. All the regression analyses were
performed using the SPSS 14.0 software package for
Windows, while for AP we calculated the Geary spatial
autocorrelation index using the ADE-4 free software for
Windows (Thioulouse et al., 1997).
Ecological analysis
In this part of the work we checked for the occurrence
of different patterns of ecological constraints (biotic or
physical) across the regional/global range of S. spinosum.
The study sites were categorized as undisturbed or
disturbed, and a list of the vascular plants occurring in
each of them was compiled. The names of the species
follow Conti et al. (2005). For each site we derived
the life form spectrum from the floristic inventory; we
then related the inter-site variation in terms of life form
frequency to the ecological/management conditions
characterizing the stands.
Israel Journal of Plant Sciences 55
2007
Based on the data set (period ~1950–2000; spatial
resolution = 2.5 arc-min) described in Hijmans et al.
(2005) and available at http://www.worldclim.org/current.htm, we carried out a climate analysis throughout
the whole range of S. spinosum. In our analysis we
considered three climatic variables: the average summer rainfall (SR), the average monthly minimum temperature for the winter period, and the average monthly
maximum temperature for the summer. These variables
were chosen because they impose important constraints
on plant life in Mediterranean-type environments (Mitrakos, 1980, 1982). In accordance with the resolution of
the data set used, the values for the cited variables were
related to 5 ´ 5 km cells in a grid covering the Mediterranean basin. As a first step we investigated the climatic relationships between the geographic areas where
S. spinosum was present at least historically. Across the
species’ range we distinguished 10 geographic zones,
as illustrated in Fig. 2. We then classified such zones
through cluster analysis of the climatic variables, by
adopting Euclidean distance as similarity measure and
between-group linkage as agglomerative method.
Moreover, by assuming the Middle East as the
ecological optimum for the species, we used the related
climatic data as a training data set in a further similarity
analysis along the Mediterranean coast. In this analysis
the similarity between each cell and the training data set
was calculated in the form of Mahalanobis distances.
Afterwards, we rescaled Mahalanobis distances in
chi-square values to facilitate their interpretation. All
the geographic analyses were supported by ArcGis 9.0
ESRI software.
Evaluating the conservation status at regional level
For the risk assessment at the regional scale we followed
the IUCN protocol (IUCN, 2001) and the most recent
guidelines for its application (IUCN, 2006).
In particular, we applied two IUCN criteria (A–B)
by estimating trends in the Extent of Occurrence
(EOO) and Area of Occupancy (AOO) over the period
1950–2006. This temporal extent roughly corresponds
to 3 generations for S. spinosum. We excluded older data
(i.e., extinction events before 1950) from the EOO and
AOO estimates. Both the EOO and AOO analyses were
based on a geo-database (including literature/herbarium
records as well as new records from our field surveys)
illustrating current species distribution in Italy and the
situation before 1950. In estimating EOO, firstly we
drew a minimum convex polygon including all the sites
of occurrence at a given time, from which we then excluded large portions of obviously unsuitable areas by
deriving the correspondent α-hull through the Delauney
triangulation (Burgman and Fox, 2003). In the AOO
Fig. 2. Geographic regions with occurrence of S. spinosum considered in the cluster analysis.
estimates, we used a 2 ´ 2 km grid for the whole Italian
range of S. spinosum with the exclusion of Sicily. In this
region the plant occurs in many sites that are close to
one another, and the more detailed information was fitted to a 10 ´ 10 km grid. However, as suggested by the
IUCN (2006) and Kunin (1998), we rescaled such estimates of AOO at broader scale, by calculating the scale
correction factor (c). From the calculations we have
obtained c = 0.66, but we upgraded this value to 0.7
because of the low dispersal ability (due to isolation),
habitat specificity, and the reduced habitat availability
of the species in Italy.
Finally, we analyzed the trends in suitable habitats
for S. spinosum over the last few decades (from 1950 to
1990). The habitat changes were evaluated in an area of
50 km2 located in northeastern Calabria (Fig. 1), which
we considered to be representative for the overall condition of S. spinosum in Italy. We used two series of aerial
photographs, taken in 1950 and 1990, respectively. In
this way we created two models of land use (in grid
format) with a cell size of 10 ´ 10 m. Using ArcGis
9.0 and the free software Fragstat 3.3 (McGarigal and
Marks, 1995), the two land cover models were used to
evaluate quantitative/qualitative habitat changes over
the considered period.
Results
Demography
The whole set of taproot sections (Fig. 3) revealed age
classes ranging between 1 and 24 years; the young
classes are dominant, as demonstrated by the lower
frequencies evident beyond an age threshold of 5–8
years. The oldest cushion was recorded in CA (Table 1).
On average, the independent ramets were 7.8 years old.
However, we found clear differences between CA and
SI, the mean age of their ramets being 9.6 and 5.4 years,
respectively (Table 1). Such a difference is confirmed by
the age structure of the sampled populations (Fig. 4). In
CA the ramet frequency decreases in cohorts more than
8–9 years old; moreover, the younger cohorts have few
ramets, indicating reduced regeneration. In contrast, the
SI sample shows a predominance of young cohorts, as
demonstrated by the decline of the ramet frequency in
Gargano et al. / Conservation ecology of S. spinosum W-populations
Fig. 3. Age distribution within the whole sample of taproot
sections, including both independent and dependent ramets
collected in CA and SI. SD = Standard deviation.
the cohorts more than 4 years old. Such a pattern, together with the age recorded for the oldest cushion (Table 1),
reflects the recent origin of the SI population. From
the Ca and SI samples, a fitted linear regression model
showed a high correlation between cushion age and both
canopy size (r = 0.904; N = 8) and root cross-section diameter (r = 0.907; N = 8). From the former relationship
we estimated for the shrubs of AP a maximum age of 21
years. Moreover, in this stand, the Geary test indicated
a significant spatial structure in relation to cushion age
(C = 0.768; p ≤ 0.0001), with the cushion age decreasing
towards the shoreline; however, such relationship does
not appear linear, as a cubic regression model fits better
(r = 0.533; N = 48). The fitted regression curve (Fig. 5)
suggests the vegetation structure characterizing AP may
be the main factor responsible for the spatial variation
of cushion age; indeed, by discontinuing the Mediterranean maquis, the presence of salt ponds facilitates the
establishment of young S. spinosum plants.
The number of dead units was significantly related
to the amount of RAMs and SRAMs in the cushion
(r = 0.489; p = 0.01). The fitted linear regression model
for ramet age vs. ramet thickness (r = 0.919; N = 350)
allowed us to estimate a mean age of 6.5 and 3.7 years
for the dead units of CA and SI, respectively; such estimates agree with the frequencies of age cohorts shown
in Fig. 4. Considering further differences in terms of
maximum cushion age and mean age of independent
ramets (Table 1), such results suggest that in the undisturbed Italian populations the ramets died earlier than in
the Middle Eastern ones.
Ecological analysis
Based on the climate data set, in AP and Ca the summer precipitation is twice as high as SI; this may be
important in contributing to the floristic differences recognized between the study sites (Table 2). Indeed, the
frequency of phanerophytes reflects the dominance of
Mediterranean maquis species (e.g., Pistacia lentiscus
Fig. 4. Age distribution of independent ramets collected in SI (top) and CA (bottom). Overall, mean age of independent ramet
was 7.85 ± 4.7 (N = 48). SD = Standard deviation.
Israel Journal of Plant Sciences 55
2007
Table 1
Comparison of demographic parameters in Italian and Middle Eastern populations; data for latter were inferred from Seligman
and Henkin (2002)
Location
Disturbance
Number of plots
Maximum age recorded
Number of ramets/m2
Mean age of independent ramets
Number of dead ramets/m2
Mean age of dead ramets
Number of seedlings/m2
Israel
CA
AP
SI
Absent
//
34
ranging from 23 to 31.8
9.80
9.4
between 7 and 12
none
Absent
4
24
7.0 ± 5.3
9.56 ± 5.2
5.0 ± 2.8
6.48 ± 0.8*
0.75 ± 1.5
Absent
//
21*
//
//
//
//
//
Fire, Grazing
4
12
5.0 ± 3.6
5.45 ± 2.4
4.5 ±3.4
3.69 ± 1.0*
1.25 ± 2.5
*—values estimated from regression models.
Table 2
Climate characteristics and life-forms’ abundance (%) in the
study sites
TMS
TmW
PS
Chamephytes
Geophytes
Hemicriptophytes
Nanophanerophytes
Phanerophytes
Therophytes
Fig. 5. Variation of estimated cushion age along the transect in
AP. SL = Shoreline.
L., Calicotome infesta (C. Presl) Guss., Cistus monspeliensis L., and Myrtus communis L.) in the vegetation
cover in AP. In contrast, CA and SI are mainly characterized by herbaceous vegetation, the woody plants
being gradually replaced by nanophanerophytes (e.g.,
S. spinosum, Daphne gnidium L., Euphorbia characias
L., Phlomis fruticosa L.) and chamephytes (Dorycnium
hirsutum (L.) Ser., Ononis natrix L., Teucrium polium
L.) with a cushion-like habitus (Table 2). The management regime gives further evidence for such differences.
Specifically, the combination of summer drought, fire,
and grazing plays a major role in limiting woody vegetation in SI and in favoring active expansion and pioneering on the part of young S. spinosum populations (Fig. 4
and Table 1). In contrast, in AP, the absence of grazing
and other land uses in a protected area that is otherwise
AP
CA
SI
29.9
6.5
55
9.1
9.1
27.3
9.1
27.3
9.1
29.5
7.7
55
11.6
7.0
39.5
14.0
14.0
14.0
26.3
4.9
22
16.7
26.7
30.0
13.3
3.3
10.0
Abbreviations and units: TMS—average of the monthly
maximum summer temperatures (°C); TmW—average of the
monthly minimum winter temperatures (°C); PS—summer
precipitation (mm).
unmanaged favors the restoration of the maquis, thus
limiting S. spinosum to the coastline (Fig. 5).
Different climatic patterns also characterize the
range of S. spinosum, and they appear to affect its
conservation status at regional level. The descriptive
statistics in Table 3 suggest that the greatest variation
occurs between the Balkan coast and the Middle East.
Further differences (summer temperatures and rainfall)
are found between the Italian peninsula and Sardinia/
Sicily. Such patterns are highlighted by the multivariate
analysis. The cluster analysis divided the regions with
occurrence of S. spinosum into three groups (Fig. 6) by
highlighting the bioclimatic affinities between regions
where the species is abundant (Middle East and southern Mediterranean coast) and separating the areas where
it becomes rare (Balkan coast and Italian peninsula).
Gargano et al. / Conservation ecology of S. spinosum W-populations
Table 3
Mean climatic data (± standard deviation) relative to winter and summer months for different regions in the S. spinosum range.
Basic data are derived from Hijmans et al. (2005); mean values were calculated on grid cells included in each considered geographic area
Areas
Mean summer temperatures (°C)
Palestine
Israel
Jordan
Cyprus
Syria
Lebanon
Greece
Turkey
Sicily
Sardinia
Tunisia
S Balkan coast
C Balkan coast
N Balkan coast
Italian peninsula
Mean winter temperatures (°C)
32.4 ± 2.9
32.3 ± 2.2
32.0 ± 2.7
31.7 ± 1.6
31.8 ± 2.9
27.7 ± 3.2
28.7 ± 2.2
29.8 ± 2.5
27.1 ± 1.5
26.9 ± 1.2
32.5 ± 1.5
26.0 ± 3.1
25.6 ± 2.5
25.0 ± 2.6
26.4 ± 2.3
8.0 ± 1.5
8.1 ± 1.2
4.7 ± 2.5
6.1 ± 1.4
3.0 ± 2.1
3.3 ± 4.2
3.4 ± 3.2
2.4 ± 3.0
6.3 ± 1.9
5.6 ± 1.4
6.1 ± 1.1
0.7 ± 2.0
1.1 ± 2.2
0.6 ± 2.8
2.8 ± 2.4
Fig. 6. Dendrogram produced by cluster analysis of climatic
data (minimum winter temperatures, maximum summer
temperatures, summer rains) for different areas of S. spinosum range. Similarity measure = Euclidean distance; Agglomerative method = Between-group linkage.
Therefore, by comparing the results of the spatial analysis performed on climatic data with the characteristic of
the species range (Fig. 7), we can infer the occurrence of
higher frequency of extinction and increasing isolation
under less suitable climatic conditions.
Evaluating the conservation status at regional level
During recent decades a series of local extinctions have
been recorded in western populations of S. spinosum
(Table 4). Based on the literature, herbarium data, and
Israel Journal of Plant Sciences 55
2007
Mean summer rainfalls (mm)
0.0 ± 0.0
0.0 ± 0.0
0.0 ± 0.0
2.7 ± 2.2
1.1 ± 1.9
0.6 ± 0.6
17.9 ± 10.8
12.0 ± 5.9
9.3 ± 3.6
12.9 ± 4.2
5.3 ± 4.0
60.7 ± 5.7
60.6 ± 7.5
69.9 ± 18.1
41.5 ± 19.9
field surveys, some populations disappeared in the late
19th/early 20th centuries, but more than 60% of extinctions have occurred since 1950 (Table 4). In Italy, such
events have significantly affected regional trends of
EOO and AOO in the period 1950–2006. Calculations
from α-hulls relative to 1950 and 2006 (Fig. 8) give an
EOO estimate of 64,412 and 30,412 km2, respectively,
indicating a decline of 52.78%. The observed decline of
the EOO of S. spinosum (whose causes have not ceased,
see below) allows for an “Endangered” status under the
IUCN’s A2 criterion. Based on the number of occupied
cells in a grid (cell size 2 ´ 2 km for Sardinia and Italian
Peninsula and 10 ´ 10 km for data from Sicily), the AOO
estimates (Table 5) are 1,684 and 1,656 km2 for the same
period; these amounts are congruent with a “Vulnerable”
status under the IUCN’s B2 criterion. However, by applying a scale correction factor to the data from Sicily,
we estimate that the Italian AOO has declined from 252,1
km2 (before 1950) to 224,1 km2 (present day). AOO in
Italy has thus fallen by 11.1% during the considered period; currently it reaches the threshold for “Endangered”
status under the IUCN’s B2 criterion.
The analysis of habitat changes give further insights
into the conservation trends of S. spinosum in Italy, suggesting that the reasons for its distributive trends may be
related to recent habitat changes. Indeed, over the period
1950–1990, the species’ habitat has been subjected to
a quantitative and qualitative erosion. As illustrated in
Table 6, the fraction of landscape covered by garigues,
the most suitable habitat for the species, has declined
Fig. 7. Global distribution of S. spinosum showing variation of species abundance across its range (top), and analysis of key
climatic parameters (bottom); dark coloration in latter indicates high similarity with Middle Eastern areas of S. spinosum
range.
in NE Calabria by 29% over the last 40 years. Habitat
loss has been accompanied by overall size reduction and
increasing isolation at patch level (see Largest Patch
and Proximity indexes in Table 6). This may be related
to the increasing landscape heterogeneity indicated by
the Shannon index values, which increased from 1.513
in 1950 to 1.551 in 1990. In completing the risk assessment for S. spinosum, it is relevant that the Italian
Gargano et al. / Conservation ecology of S. spinosum W-populations
10
Country
Croatia
Italy
Italy
Italy
Italy
Italy
Italy
Italy
Table 4
Exctint populations of S. spinosum in the W-Mediterranean basin
Locality
Cause of extinction
Near Fiume and Buccari
Not specified
Tivoli (Latium)
Anthropic
Bari (Apulia)
Not specified
Between Mola and Bari (Apulia)
Not specified
Torre Colimena (Apulia)
Anthropic
Penisola della Strea (Apulia)
Anthropic
Capo Colonna (Calabria)
Anthropic
Stagno di Santa Gilla-Assemini (Sardinia)
Anthropic
Last recorded
Before 1900
1900–1950
Before 1900
Before 1900
After 1950
After 1950
After 1950
After 1950
Source
Martinoli (1969)
Pignatti (1982)
Caniglia et al. (1974)
Caniglia et al. (1974)
Caniglia et al. (1974)
Unpublished data
Unpublished data
Unpublished data
Table 5
Area of Occupancy (AOO) estimates, before and after local extinctions of Italian S. spinosum populations
Grid scale (km)
Region
2 ´ 2
2 ´ 2
2 ´ 2
2 ´ 2
10 ´ 10
Apulia
Calabria
Latium
Sardinia
Sicily
Occupied cells
(after 1950)
4
14
1
2
16
Occupied cells
(before 1950)
1
12
0
1
16
AOO before
1950 (km2)
16
56
4
8
1600
Current AOO
(km2)
4
48
0
4
1600
Table 6
Parameters describing habitat trends for S. spinosum in NE
Calabria over the period 1950–1990
Year
Area (% of
landscape) LPI
PROX
1950
1990
1950
1990
1950
1990
Habitat type
Garigues on
fixed sediments
Coastal
garigues
0.78
0.29
0.69
0.11
3.8
2.39
38.21
27.02
1.71
0.64
9.95
6.52
Abbreviations: LPI—largest patch index; PROX—proximity
index.
Fig. 8. Local extinctions of S. spinosum in Italy (A) and their
effects on regional Extent of Occurrence (B). A: crosses
identify extinct populations, triangles correspond to current
ones. B: gray background shows extent of convex hull in
1950 before Delauney triangulation; dashed lines delimit
lost portion of α-hull after 1950, continuous lines represent
current (2006) EOO, which includes a disjunct population in
Sardinia.
Israel Journal of Plant Sciences 55
2007
S. spinosum population appears to have little chances
for recruitment from the global population because of its
isolation; in addition most of the Italian subpopulations
are themselves isolated from one another. Therefore,
the IUCN procedures allow the regional population
of S. spinosum to qualify as “Endangered” (EN): A2c;
B2ab(ii,iii).
Discussion
According to Seligman and Henkin (2002), the independent ramets are older than the mean age we observed in
the whole set of sections (Figs. 3,4), because the clonal
11
units become independent late, when the mother is too
aged to sustain them. The age structure analysis (Fig. 4)
indicates that SI represents a younger population; it suggests a more recent spread of S. spinosum in this stand
and is congruent with an expansion dynamic within
disturbed sites, as already supposed for the Sicilian
populations (Barbagallo et al., 1979). Considering the
life history strategy of S. spinosum, it is significant that,
in the studied populations, the ramet density was lower
than in the Middle Eastern ones (Table 1). Indeed, under
ecologically suitable conditions, the species should be
able to exclude other competitors by adopting a phalanx-type clonal strategy (Lovett Doust, 1981). At the
population level this would mean that “the canopies
can expand and coalesce to form a continuous cover
of S. spinosum over tens to hundreds of square meters”
(Seligman and Henkin, 2002). By following such a
strategy the species can become dominant across large
areas and in different habitats, as reported by Zohary
(1973) for the Middle East. In contrast, our findings indicate a reduced vegetative vigor for the species in Italy,
also evident from phytosociological data (e.g., Géhu et
al., 1984; Biondi and Mossa, 1992; Brullo et al., 1997),
since, at the community level, the discontinuous cover
of S. spinosum allows other species to establish themselves with comparable frequencies. Such a reduced
density of clones may be the consequence either of a
lower net production or a shorter ramet lifespan. On
average, in Middle Eastern areas the expected life span
for the ramets of S. spinosum appears to be less than
15 years, and individuals more than 20 years old are
very rare, even in long-undisturbed stands (Litav and
Orshan, 1971; Seligman and Henkin, 2002). However,
the age of the oldest ramet, together with the mean age
of the dead ramets (Table 1) and the lack of ramets in
age classes older than 9–10 years (Fig. 3), provides
evidence that in the undisturbed Italian populations the
ramets die earlier. Furthermore, it should be noted that
the sampling protocol allows for the identification of
the dead ramets (Seligman and Henkin, 2002); nevertheless, even adding the dead ramets to the live ones,
in Italy their density remains much lower than in the
Middle Eastern populations. This occurrence suggests
lower productions of clones per individual in the studied populations of S. spinosum. Since adverse climatic
conditions have often been invoked as a constraint for
the border populations (Thomas et al., 2001; Holt and
Keitt, 2005), they may be responsible for the differences
discussed above. Indeed, the climate of the study area
is very different from that of the eastern portion of the
species’ range (Figs. 6,7); specifically, Table 3 indicates
a marked decrease in the summer drought conditions
westwards. Therefore, higher availability of moisture
during the summer is likely to favor the woody species
of the Mediterranean maquis as Pistacia lentiscus and
Calicotome infesta, confining S. spinosum to stands
that are edaphically unsuitable for them. Accordingly,
phytosociological studies suggest that the Italian communities with S. spinosum occur in habitats too poor for
the establishment of dense Mediterranean shrublands
(Barbagallo et al., 1979; Géhu et al., 1984; Bartolo et
al., 1986; Biondi and Mossa, 1992; Biondi et al., 1994;
Brullo et al., 1997). Hence, in southeastern Sicily reduced summer moisture in comparison to the Italian
peninsula (Table 3 and Figs. 6,7) may be significant
in improving the status of S. spinosum by reducing the
chances for rich maquis vegetation. This agrees with our
analysis of the growth forms frequency in the study sites
(Table 2), and suggests possible explanations for the
spatial structure detected in AP. With regard to this, it
should be mentioned that in S. spinosum populations the
occurrence of age spatial patterns over short distances
is likely. Reisman-Berman et al. (2006) found that in
S. spinosum significant spatial structures of age distribution are a consequence of dispersal limitation and
reduced establishment rates (due to resource limitation
produced by the intra-specific competition). Likewise,
Malkinson and Jeltsch (2007) point out the role of intraspecific competition in generating spatial patterns in the
populations of S. spinosum under conditions of increasing moisture availability. In contrast, our findings may
reflect the contribution of inter-specific competition in
segregating older plants in patches with dense cover of
Mediterranean maquis. Therefore, in AP an important
role seems to be played by the dynamics of maquis restoration in forcing S. spinosum to colonize areas unsuitable for (or not yet occupied by) other woody species.
The limited importance of sexual recruitment may
strengthen the influence of environmental conditions
on the spread and survival rates of S. spinosum. Indeed,
although high germination rates contribute to a rapid
recovery after disturbance through massive seedling
production (Roy and Arianoutsou-Faraggitaki, 1985;
Seligman and Henkin, 2000), subsequently sexual recruitment becomes marginal for the species. Hence, as
also suggested by the low frequency of seedlings that
often occur in S. spinosum populations, we can assume
clonal sprouting as the main strategy for the expansion
and persistence of the species (Seligman and Henkin,
2002). On the other hand, this may reduce ecological
flexibility under variable or unsuitable environments,
accounting for the importance of ecological limitations
to the conservation status of the species at the edges of
its range. Indeed, as highlighted by Channell and Lomolino (2000) and Ezard and Travis (2006), habitat confinement due to unsuitable conditions may be respon-
Gargano et al. / Conservation ecology of S. spinosum W-populations
12
sible for increasing isolation in peripheral populations;
consequently, they suffer higher demographic turnover
and extinction rates (e.g., Nantel and Gagnon, 1999;
Lienert et al., 2002). Such considerations account for the
westward fragmentation of S. spinosum distribution and
the frequent extinctions documented for Dalmatia, Tunisia, and Italy (Martinoli, 1969; Caniglia et al., 1974;
Pignatti, 1982). As we have highlighted for South Italy,
human-induced habitat loss can exacerbate the threat to
S. spinosum’s western populations because it is unable
to colonize new stands. The consequence is a rapid decline in Extent of Occurrence and Area of Occupancy,
as we have found for the Italian population.
Our work provides a further example of the ecological constraints affecting border populations of widespread species. The combination of these limitations,
together with the species’ own biological traits (clonal
strategy) and the patterns of human disturbance (i.e.,
habitat loss and, in some cases, the restoration dynamics
of the vegetation), can contribute to increasing levels of
risk. In such conditions, habitat specificity can become
the ecological base for a status of regionally endangered
species. In Italy as well as in most S. spinosum western
sites, adequate in/ex situ conservations strategies are required in order to avoid further species decline. Habitat
preservation as well as control of vegetation dynamics
could be important for in situ conservation. In addition,
since good seed viability has also been found in isolated
Italian populations (Tornadore Marchiori et al., 1977),
ex situ conservation strategies should include seed banking for future reintroduction planning, as already carried
out by the Sardinia Germoplasm Bank (BG-SAR).
Acknowledgments
The authors are grateful to Prof. Graziano Rossi (University of Pavia), Prof. Gianluigi Bacchetta (University
of Cagliari), and Prof. Aldo Musacchio (University of
Calabria) for the valuable comments on earlier drafts
of the manuscript, and to George Metcalf (University
of Salento) for linguistic improvement. Prof. Salvatore
Brullo (University of Catania) and Dr. Joseph Buhagiar
(University of Malta) are thanked for providing unpublished data. Finally, we also thank Dr. Lorenzo Peruzzi
(University of Pisa) for his support in plant identification, and Dr. Katia Caparelli and Dr. Gabriella Aquaro
for their help in the field. The research the manuscript
refers to is a part of the study program known as “Iniziativa per l’implementazione in Italia delle categorie e
dei criteri IUCN (2001), per la redazione di nuove liste
rosse” promoted by Società Botanica Italiana (Gruppi
di Lavoro per la Conservazione della Natura, Floristica
e Micologia).
Israel Journal of Plant Sciences 55
2007
References
Barbagallo, C., Brullo, S., Fagotto, F. 1979. Boschi di Quercus
ilex L. del territorio di Siracusa e principali aspetti di degradazione. Pubbl. Ist. Bot. Univ. Catania, 2.
Bartolo, G., Brullo, S., Minissale, P., Spampinato, G. 1985.
Osservazioni fitosociologiche sulle pinete a Pinus halepensis Miller del bacino del fiume Tellaro (Sicilia Sud-orientale). Boll. Acc. Gioenia Sci. Nat. 18(325): 255–270.
Biondi, E., Mossa, L. 1992. Studio fitosociologico del promontorio di Capo S. Elia e dei Colli di Cagliari (Sardegna).
Doc. Phytosoc. 14: 1–44.
Biondi, E., Ballelli, S., Allegrezza, M., Taffetani, F., Francalancia, C. 1994. La vegetazione delle “fiumare” del versante ionico lucano-calabrese. Fitosociologia 27: 51–66.
Brullo, S., Minissale, P., Spampinato, G. 1997. La classe Cisto-Micromerietea nel Mediterraneo centrale ed orientale.
Fitosociologia 32: 29–60.
Burgman, M.A., Fox, J.C. 2003. Bias in species range estimates from minimum convex polygons: implications for
conservation and options for improved planning. Anim.
Conserv. 6: 19–28.
Busch, J.W. 2005. The evolution of self-compatibility in
geographically peripheral populations of Leavenworthia
alabamica (Brassicaceae). Am. J. Bot. 92: 1503–1512.
Caniglia, G., Chiesura Lorenzoni, F., Curti, L., Lorenzoni,
G.G., Marchiori, S. 1974. Sarcopoterium spinosum (L.)
Spach ed Anthyllis hermanniae L., due relitti floristici nel
Salento (Puglia). Iv Simposio nazionale sulla conservazione della natura I: 315–330.
Channell, R., Lomolino, M.V. 2000. Dynamic biogeography and
conservation of endangered species. Nature 403: 84–86.
Chiesura Lorenzoni, F., Lorenzoni, G.G. 1977. Significato
fitogeografico e fitosociologico delle cenosi a Sarcopoterium spinosum (L.) Spach di Capo S. Elia (Cagliari–Sardegna Meridionale). Giorn. Bot. Ital. 111: 263–276.
Conti, F., Abbate, G., Alessandrini, A., Blasi, C., eds. 2005. An
annotated checklist of the Italian vascular flora. Palombi
Editori, Rome.
Ezard, T.H.G., Travis, J.M.J. 2006. The impact of habitat loss
and fragmentation on genetic drift and fixation time. Oikos
114: 367–375.
Géhu, J.M., Costa, M., Scoppola, A., Biondi, E., Marchiori,
S., Peris, J.B., Franck, J., Caniglia, G., Veri, L. 1984. Essai
synsystematique et synchorologique sur les végétations
littorales italiennes dans un but conservatoire. Doc. Phytosoc. 8: 393–421.
Grant, M.C., Antonovics, J. 1978. Biology of ecologically
marginal populations of Anthoxanthum odoratum. I. Phenetics and dynamics. Evolution 32: 822–838.
Herlihy, C.R., Eckert, C.G. 2005. Evolution of self-fertilization at geographical range margins? A comparison of
demographic, floral, and mating system variables in central
vs. peripheral populations of Aquilegia canadensis (Ranunculaceae). Am. J. Bot. 92(4): 744–751.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis,
A. 2005. Very high resolution interpolated climatic surfaces for global land areas. Int. J. Climatol. 25: 1965–1978.
13
Holt, R.D., Keitt, T.H. 2005. Species’ borders: an unifying
theme in ecology. Oikos 108: 3–6.
IUCN, 2001. IUCN Red List Categories and Criteria. Version
3.1. IUCN-SSC, Gland, Switzerland and Cambridge, UK.
IUCN, 2006. Guidelines for using IUCN Red List Categories
and Criteria. IUCN-SSC, Gland, Switzerland and Cambridge, UK.
Jonas, C.S., Geber, M.A. 1999. Variation among populations
of Clarkia unguiculata (Onagraceae) along altitudinal and
latitudinal gradients. Am. J. Bot. 86: 333–343.
Kunin, W.E. 1998. Extrapolating species abundances across
spatial scales. Science 281: 1513–1515.
Lammi, A., Siikamäki, P., Mustajärvi, K. 1999. Genetic diversity, population size, and fitness in central and peripheral
populations of a rare plant Lychnis viscaria. Conserv. Biol.
13: 1069–1078.
Lavrentiades, G.J. 1969. Studies on the flora and vegetation
of the Ormos Archangelou in Rhodos Island. Vegetatio
19(1–6): 308–329.
Lesica, P., Allendorf, F.W. 1995. When are peripheral
populations valuable for conservation? Conserv. Biol. 9:
753–760.
Lienert, J., Fischer, M., Diemer, M. 2002. Local extinctions of
the wetland specialist Swertia perennis L. (Gentianaceae)
in Switzerland: a revisitation study based on herbarium
records. Biol. Conserv. 103: 65–76.
Litav, M., Orshan, G. 1971. Biological flora of Israel. 1. Sarcopoterium spinosum (L.) Spach. Isr. J. Bot. 20: 48–64.
Lönn, M., Prentice, H.C. 2002. Gene diversity and demographic turnover in central and peripheral populations
of the perennial herb Gypsophila fastigiata. Oikos 99:
489–498.
Lovett Doust, L. 1981. Population dynamics and local specialization in a clonal perennial Ranunculus repens. I. The
dynamics of ramets in contrasting habitats. J. Ecol. 69:
743–755.
Malkinson, D., Jeltsch, F. 2007. Intraspecific facilitation: a
missing process along increasing stress gradients – insights from simulated shrub populations. Ecography 30(3):
339–348.
Martinoli, G. 1969. Poterium spinosum L. elemento del Mediterraneo Orientale a Calamosca (Capo S. Elia, Sardegna).
Giorn. Bot. Ital. 103: 325–340.
McGarigal, K., Marks, B.J. 1995. FRAGSTATS: spatial pattern analysis program for quantifying landscape structure.
USDA Forest Service General Technical Report PNW351.
Médail, F., Ziman, S., Boscaiu, M., Riera, J., Lambrou, M.,
Vela, E., Dutton, B., Ehrendorfer, F. 2002. Comparative
analysis of biological and ecological differentiation of
Anemone palmata L. (Ranunculaceae) in the western Mediterranean (France and Spain): an assessment of rarity and
population persistence. Bot. J. Linn. Soc. 140: 95–114.
Mitrakos, K. 1980. A theory for Mediterranean plant life. Acta
Oecol.-Oec. Plant. 1(15): 245–252.
Mitrakos, K. 1982. Winter low temperatures in mediterraneantype ecosystems. Ecol. Medit. 8(1–2): 95–102.
Moritz, C. 2002. Strategies to protect biological diversity and
the evolutionary processes that sustain it. Syst. Biol. 51:
238–254.
Nantel, P., Gagnon, D. 1999. Variability in the dynamics of
northern peripheral versus southern populations of two
clonal plant species, Helianthus divaricatus and Rhus aromatica. J. Ecol. 87: 748–760.
Pignatti, S. 1982. Flora d’Italia 1. Edagricole, Bologna.
Proctor, M.C.F. 1968. Sarcopoterium Spach. In: Tutin, T.G.,
Heywood, V.H., Burges, N.A., Moore, D.M., Valentine,
D.H., Walters, S.M., Webb, D.A., eds. Flora Europaea 2.
Cambridge University Press, Cambridge, p. 34.
Reisman-Berman, O., Kadmon, R., Shachak, M. 2006. Spatiotemporal scales of dispersal limitation in the recolonization
of a semi-arid Mediterranean old-field. Ecography 29:
418–426.
Roy, J., Arianoutsou-Faraggitaki, M. 1985. Light quality as the
environmental trigger for the germination of the fire-promoted species Sarcopoterium spinosum L. Flora 177: 345–349.
Sagarin, R.D., Gaines, S.D. 2002. The ‘abundant centre’ distribution: to what extent is it a biogeographical rule? Ecol.
Lett. 5: 137–147.
Sagarin, R.D., Gaines, S.D. 2006. Recent studies improve understanding of population dynamics across species range.
Oikos 115: 386–388.
Seligman, N., Henkin, Z. 2000. Regeneration of a dominant Mediterranean dwarf-shrub after fire. J. Veg. Sci. 11: 893–902.
Seligman, N., Henkin, Z. 2002. Persistence in Sarcopoterium spinosum dwarf-shrub communities. Plant Ecol. 164:
95–107.
Shea, K.L., Furnier, G.R. 2002. Genetic variation and population structure in central and isolated populations of balsam
fir, Abies balsamica (Pinaceae). Am. J. Bot. 89: 783–791.
Thioulouse, J., Chessel, D., Doledec, S., Olivier, J.-M. 1997.
ADE-4: a multivariate analysis and graphical display software. Stat. Comput. 7: 75–83.
Thomas, C.D., Bodsworth, E.J., Wilson, R.J., Simmons, A.D.,
Davies, Z.G., Musche, M., Conradt, L. 2001. Ecological
and evolutionary processes at expanding range margins.
Nature 411: 577–581.
Thompson, J.D. 2005. Plant evolution in the Mediterranean.
Oxford University Press, Oxford.
Tornadore Marchiori, N., Caniglia, G., Chiesura Lorenzoni, F.,
Curti, L., Lorenzoni, G.G., Marchiori, S., Razzara, S. 1977.
Sarcopoterium spinosum (L.) Spach in Puglia. Vitalità germinativa comparata. Giorn. Bot. Ital. 112: 307–308.
Volis, S., Mendlinger, S., Olsvig-Whittaker, L., Safriel, U.N.,
Orlovsky, N. 1998. Phenotypic variation and stress resistance in core and peripheral populations of Hordeum spontaneum. Biodivers. Conserv. 7: 799–813.
Volis, S., Mendlinger, S., Orlovsky, N. 2000. Variability in
phenotypic traits in core and peripheral populations of
wild barley Hordeum spontaneum Koch. Hereditas 133:
235–247.
Zohary, M. 1973. Geobotanical foundations of the Middle
East. Gustav Fisher Verlag, Stuttgart and Swets & Zeitlinger, Amsterdam.
Gargano et al. / Conservation ecology of S. spinosum W-populations