Page 10 - v14n2a04

Fattorini & Dapporto
Revista de Gestão Costeira Integrada / Journal of Integrated Coastal Zone Management 14(2):185-197 (2014)
composition. The BCC1 highlighted the conservation
importance of very remote, although faunistically poor,
islands, a result also supported by the BCW1 index. These
remote islands host few species as a consequence of their
distance from possible source areas, yet these species have
a narrow distribution, occurring on one or few islands.
Moreover, as indicated by results achieved using BCC1p and
BCW1p indices, the faunas of these islands tend to be also
characterised by species that are distributed on islands whose
summed perimeter determines high rarity scores. This may
appear surprising, because typically the most remote islands
are those colonized by animal species with high dispersal
power, which are also the most widely distributed among
islands and which should have a total distribution with high
perimeter. The fact that the most remote islands are usually
colonised by widespread species is a common phenomenon
and it is at the basis of the nested structure of most island
faunas (Dennis et al. 2012; Ulrich et al. 2009). In the case of elevation islands, but because they also have many common
the tenebrionid beetles inhabiting the Italian small islands,
two factors may have contributed to this unexpected pattern.
First, tenebrionids tend to be sedentary animals, with low
dispersal power, with most of the species found on small
islands being flightless (Fattorini, 2002, 2008a, b, 2010a).
This led to uncommon colonization events mostly favored
by past land bridges and to repeated evolution of endemic
forms (usually considered as subspecies), especially on the
most isolated islands. Second, some of the most remote
and highest priority islands are not exclusively sourced
from the Italian peninsula and major islands but received
several elements from the nearer North African mainland
(Pantelleria, Linosa, Lampione and Lampedusa). Thus, these
islands host several North African species that are absent from in the sample (Dapporto & Dennis, 2008).
Italy and, hence, from all other islands. Interestingly, while
BCC2 and BCC3 did not correlate with island distance,
BCW2 and BCW3 were correlated with island isolation.
This is consistent with the fact that endemic species, which
have high conservation weights, tend to be more numerous
on the most remote islands, thus increasing the BCW2 and
BCW3 values for remote islands.
Another important result is the negative relationships of
BCC2 and BCC3 with altitude, thus habitat diversity, which
can be explained by the ecology of most of the tenebrionid
species considered in this study. The BCC2 index included
species’ association with coastal environment as a factor of
vulnerability in contrast with a broader habitat specificity.
Although some low-elevation islands were occupied by
forests before human exploitation (e.g. Lampedusa), in
general, flat islands tend to be almost entirely occupied by
coastal environments, and hence their faunal assemblages
are mainly composed of tenebrionid species that were scored
as rare for habitat because of their strict association with
coastal environments. By contrast, islands with relatively
higher elevations tend to be occupied also by non-coastal
environments, such as the high Mediterranean maquis. The
occurrence on these islands of tenebrionid species not strictly Ustica and Pantelleria have been prioritised by one (Ustica)
associated with coastal environments determines a decrease
in the proportion of tenebrionids scored as rare for habitat
in the BCC2 index, which led to a negative relationship
between this index and island elevation.
- 194 -
The BCC3 index also had a negative correlation
with island elevation. Because this index is based on a
biogeographical categorization of species, this result implies
that higher islands host species assemblages with a lower
proportion of endemic species. This may be due to the fact
that while endemic species are always a small number, a
higher environmental diversity may promote the presence
of many non-endemic species coming from the mainland.
Because of the increase in the number of non-endemic
species, endemics will represent a lower proportion of the
tenebrionid faunas, thus leading to a decrease in the BCC3
index. Interestingly, BCW1 and BCW2 (constructed just to
cope with potential over-indexing biases for poorest islands)
were positively correlated with elevation. This indicates that
the negative correlations found for BCC2 and BCC3 may
be due to the fact that environmentally more diverse islands
may even have as many or more rare species than the low-
species in certain habitats, the average rarity is lower.
The BCC2 and BCC3 results suggests that islands may
be important not only for the presence of several endemic
species, but also as ecological refugia for species associated with
the particular coastal biotopes that have been deteriorated
on more largely populated mainland areas. It should be also
noted that BCC indices are not correlated with island area,
whereas a positive correlation with island area was found for
two BCW indices (BCW1 and BCW2). Lack of correlation
with island area, however, is not a bias for the BCC index
itself. For example, for the butterflies of the Italian islands
the islands showing the highest BCC value were the largest
The islands prioritised by the various indices varied
significantly in their geographical characteristics. In
particular, the set of islands recovered by the BCC1 index
had mean geographical characteristics different from those
of all other three BCC indices. By contrast, the four BCW
indices recovered similar island sets that were not significantly
different in their mean values for the geographical
characteristics used in this study.
Although the various indices gave different island
prioritisations, most of the islands included in the third
quartile for two or more of the three indices belong to the
same two archipelagos, the Tuscan Islands and the Pelagie
Islands (plus two very isolated islands, Pantelleria and
Ustica, which do not belong to any archipelago). Except
the Aegadian islands of Marettimo (prioritised by BCW1,
BCW2, BCW3) and Favignana (prioritised by BCW1 and
BCW2, respectively), no island belonging to the other Italian
archipelagos (Aeolian, Campane, Pontine, and Tremiti) was
prioritised by more than one index.
In spite of their low levels of endemism, the islands of
and two (Pantelleria) BCC indices, and by all BCW indices.
Despite Ustica and Pantelleria emerged recently, suggesting
that there was not sufficient time for higher endemicity
levels to develop, these islands are located quite far from
Sicily and mainland areas. This characteristic could have
favoured some relatively fast morphological differentiation
in some tenebrionid populations. Actually, one endemic
taxon (Opatrum validum marcuzzii) is known from Ustica