Page 2 - Tonelli_et_al_2016
P. 2

162 Tonelli et al.

Introduction                                                       Dapporto & Cini, 2007; Fattorini, 2009a, 2009b, 2010;
                                                                   Sfenthourakis, 1996; Dennis et al., 2000). Fattorini
Islands are among the areas of the world that have                 (2010), for example, investigated the importance of
most aroused the curiosity of naturalists, long attracted          island areas, distance from the continent, inter–island
by their peculiarities such as geographical isolation,             distance and island age of the Aeolian archipelago
small size and, at least for volcanic islands, recent              (southern Italy) in determining patterns of spatial
age (Whittaker & Fernández�–�P�a�l�a�c�io�s�,��2�0�0�7��).��B�e��  variation in beta diversity for several taxa, including
cause of the relative simplicity of studying an island             dung beetles. In this research, Fattorini concluded
(Vitousek, 2002), many studies of these environments               that the dung beetle fauna originated quite recently
have become paradigmatic (MacArthur & Wilson,                      and that the species were established on the islands
1967; Simberloff, 1974). Islands are ideal contexts                by a 'stepping stone' dispersive process: species dis�
to conduct a natural historical experiment (Diamond                persed from one island to the nearest (MacArthur &
& Robinson, 2010).                                                 Wilson, 1967). However, in an attempt to reconstruct
                                                                   the dispersal patterns of coprophagous fauna, Fattorini
   In biogeographical terms, we can distinguish two                (2010) identified conflicting results. Specifically, his
major categories of islands: continental and oceanic.              analysis of similarity (with Jaccard and Kulczynski 2
The former are located on the continental shelf and                indices) produced three different clusters, which cannot
may have been linked to the mainland in the past,                  provide an unambiguous explanation. Furthermore,
while the latter have never been connected to the                  these indices are not appropriate for the exploration
mainland (Whittaker & Fernández–Palacios, 2007).                   of dispersal fluxes because they cannot identify the
This distinction has conceptual implications of great              direction of dispersal. Using an expanded database
importance. Since the beginning of the modern                      (in terms of areas and dung beetle species) and
debate 'dispersal versus vicariance' (Croizat et al.,              inspired by Fattorini’s (2010) paper, we attempted in
1974; Heads, 2014), the dispersalist approaches have               this exploratory research to reconstruct the possible
undergone considerable criticism (Morrone & Crisci,                dispersal routes of dung beetle fauna in the coloni�
1995 and references cited therein). The non–refut�                 zation of circum–Sicilian volcanic islands in order to
ability of dispersalist explanations, defined ad hoc (for          formulate a posteriori hypothesis about the probable
review: Morrone & Crisci, 1995; Bueno Hernández &                  mechanisms involved in the conformation of these
Llorente–Bousquets, 2000; Zunino & Zullini, 2004),                 island assemblages.
has shifted focus towards vicariantist arguments.
This paradigm shift has consequently led to a lack                 Material and methods
of interest in respect to oceanic islands due to the
non–falsifiability of hypotheses (Cowie & Holland,                 Study area
2006). However, in recent years, the importance of
dispersal as a key mechanism in determining species                The survey focused on 10 volcanic islands: eight loca�
distribution has been emphasized (De Queiroz, 2005;                ted north of Sicily (Ustica and the Aeolian archipelago:
Cowie & Holland, 2006) and oceanic islands have                    Lipari, Salina, Vulcano, Stromboli, Filicudi, Alicudi,
attracted attention because their biodiversity pattern             Panarea) and two in the Sicily Channel (Pantelleria
must necessarily be explained in dispersal terms                   and Linosa) (fig. 1).
(Cowie & Holland, 2006). In this context, the value
of biodiversity analysis methods should be stressed                   At no point in the past were these islands con�
(Magurran & McGill, 2011), especially regarding the                nected to continental areas, allowing us to exclude
beta diversity that Whittaker (1960, 1972) has defined             vicariance events. The oldest islands are Linosa and
as “the extent of change in community composition”.                Ustica (about 1,000 Kyr). Their age excludes any
In recent years, studies concerning the beta diversity             involvement in the Messinian salinity crisis, which
have proliferated (Anderson et al., 2011). This inter�             ended about 5.3  million years ago (Krijgsman et al.,
est in the study of beta diversity is linked to a basic            1999). This period had a tremendous impact on the
question of community ecology —what makes a set                    biogeography of the Mediterranean fauna (Sanmartín,
of species more or less similar to another in different            2003; Marra, 2005). The estimate of the most severe
times and spaces (Vellend, 2010)?.                                 sea–level drops over the last 5.3 million years is –120
                                                                   m (Rohling et al., 2014). The Aeolian archipelago is
   Several studies on different taxa and scales have               separated from Sicily by a channel between 1,000 and
addressed the issue of the distribution, composition               2,000 m deep, while the interisland depth varies from
and species richness of organisms from an eco–bio�                 400 to 1,400 m. Only Lipari and Vulcano, divided by
geographical perspective, aiming to identify the key               water depths < 50 m, were connected to each other but
variables that could explain these patterns (Freestone             they were always separated from Sicily during glacial
& Inouye, 2006; Veech & Crist, 2007; Qian, 2009; Bin               periods of sea–level lowering. Ustica is the summit of
et al., 2010; Jiménez–Valverde et al., 2010; Vellend,              a large volcanic edifice resting on the seabed at depths
2010; Baselga et al., 2012; Dexter et al., 2012). These            of the order of 2,000 m (Ruggieri, 1973; Marani et al.,
studies have also focused on island environments and               2004). Pantelleria and Linosa were also isolated during
multiple studies have analysed the factors controlling             glacial phases (Shackleton et al., 1984).
the spatial patterns of biodiversity among many taxa
(Kadmon & Pulliam, 1993; Legakis & Kypriotakis, 1994;                 Although the hypothetical dispersal routes were in�
Palmer, 1998; Palmer et al., 1999; Garcia–Barros et al.,           vestigated only for volcanic islands, we also examined
2002; Guerrero et al., 2005; Hausdorf & Hennig, 2005;
   1   2   3   4   5   6   7