Page 3 - Fattorini_etal2015

Measuring influence of environmental heterogeneity on island beetles
cates, in practice, a reduction in heterogeneity). Alterna-
tively, one could use a specific terminology, such as “envi-
ronmental diversity”, “environmental dominance”, or “en-
vironmental evenness”, depending on what is being meas-
ured (see also Stein & Kreft 2014 for a review of the use
of various expressions).
The case study
To put in practice our observations, we present here a case
study in the field of island biogeography dealing, in par-
ticular, with the distribution of tenebrionid beetles on the
islands surrounding Sicily. Sicily is the largest Mediterra-
nean island, and it is surrounded by a number of small is-
lands, collectively known as the “circumsicilian” islands.
The biogeography of the circumsicilian islands is particu-
larly complex (e.g. Fattorini 2010, 2011a,b, and references
therein), because they vary greatly in terms of area (Mal-
ta, the largest island, has an area of 245.7 km2, but most of
the islands are below 30 km2), ecological and climatologi-
cal settings, geological origin (volcanic vs. sedimentary),
paleogeography (some were connected to Sicily or Africa
during Pleistocene glaciations, others remained isolated)
and distance to two different main sources of colonization
(Sicily vs. Africa) showing high degree of biotic diversi-
fication. Thus, they represent a good model to explore the
influence of many possible abiotic determinants on animal
communities including environmental heterogeneity. In
general, the fauna of these islands is also relatively well
known. Among insects, one of the best investigated groups
is the beetle family of Tenebrionidae. Thanks to their low
dispersal ability, tenebrionids represent excellent bioge-
ographical markers of historical processes and they have
been repeatedly used to investigate the biogeography of
Mediterranean islands (e.g., Fattorini 2002a,b, 2006a,b,c,
2007a,b,c,d, 2009a,b, 2010, 2011a,b, Fattorini & Fowles
2005, Fattorini et al. 1999, Hausdorf & Hennig 2005, Tri-
chas et al. 2008, Papadopoulou et al. 2008, 2009, 2010),
including the circumsicilan ones (Fattorini 2010, 2011a,b).
Material and methods
Study area
We used data from 25 circumsicilian islands, including:
the Aeolian Islands (volcanic: Stromboli, Panarea, Vulca-
no, Lipari, Salina, Filicudi, Alicudi, and seven smaller is-
lets), the Egadi Islands (sedimentary: Levanzo, Favignana,
Marettimo), the Pelagie (Linosa, volcanic, and Lampedu-
sa and Lampione, sedimentary), and the Maltese Islands
(Malta, Gozo, Comino, Cominotto, and Filfla, sedimenta-
ry); Ustica and Pantelleria (both volcanic), are rather iso-
lated. For detailed information on these islands see Maz-
zola et al. (2001), Fattorini (2010, 2011a,b) and Savona
Ventura (2011). Data on tenebrionid distribution in the cir-
cumsicilian islands were extracted from Fattorini (2011b)
and updated with new data reported by Lo Cascio & Pas-
3
ta (2012) and Lillig et al. (2012a,b). Species presence/ab-
sence on individual islands is given in Appendix 1.
Environmental categorization
For an ecological characterization of island landscapes, we
calculated the extent of island surface occupied by differ-
ent land cover categories according to the European Co-
rine Land Cover classification. Although somewhat crude
and of limited value for small regions, Corine land cov-
er categories are extensively used to express species-en-
vironment relationships (e.g., Lobo & Martín-Piera 2002,
Stefanescu et al. 2004, Maes et al. 2005). A total of 24 land
cover categories were found on the circumsicilian islands
(Italian Ministry of the Environment and Protection of the
Territory and Sea 2009, Malta Environment and Planning
Authority 2009). This number of environmental catego-
ries is disproportionately large in respect to the small num-
ber of islands composing the archipelago. Moreover, some
categories are represented by very small patches and can
be easily combined into broader categories. Thus, we used
the following broader land cover categories: Built up ar-
eas (including Continuous urban fabric, Discontinuous
urban fabric, Industrial or commercial units, Port area,
Airports, Mineral extraction sites, Dump Sites, Green ur-
ban areas, and Sports and Leisure Facilities), Cultivations
(including Vineyards, Non-irrigated arable land, Natu-
ral grassland [because most often found in areas where
there is extensive agricultural activity], Annual crops as-
sociated with permanent crops, Complex cultivation pat-
terns, and Land principally occupied by agriculture, with
significant areas of natural vegetation), Coniferous forest,
Broad-leaved and Mixed Forests, Sclerophyllous vegeta-
tion, Bare rock and Sparsely vegetated areas, and Wet ar-
eas (including Salt Marshes, Salines, and Water bodies).
Even if these broad land cover units are coarse in compari-
son to the scale at which insects perceive small-scale envi-
ronmental heterogeneity, they correspond well to distinct
keystone structures (sensu Tews et al. 2004) for tenebrio-
nid species. In particular, each of the seven classes used
here corresponded to different microclimate conditions,
food resources, and soil characteristics, which are among
the most important factors shaping tenebrionid communi-
ties in the Mediterranean (Fattorini 2008, 2009c).
Environmental heterogeneity measures
In addition to score the number of land cover categories
(N) occurring on each island, we used the land category
information to compute various synthetic indices of envi-
ronmental heterogeneity (Table 1). For this, we applied in-
dices of diversity, evenness and dominance derived from
those used in community ecology (Legendre & Legendre
1998, Magurann 1988, 2004, Hayek & Buzas 2010):
- Simpson dominance index: 2
 C =∑
Ai
A
where Ai is the extent of the land cover category i, and