Page 8 - mamm.73.3.169

Article in press - uncorrected proof
176 M. Masseti: Homogenisation and the loss of biodiversity of mammals of the Mediterranean islands
bush elephant, and hence probably a different species
(Roca et al. 2001, Debruyne et al. 2003, cf. Kingdon
2004). Not all authorities, however – notably the IUCN
African Elephant Specialist Group – consider currently
available evidence sufficient for splitting the African ele-
phant into two species. More specifically, the disputed
pigmy elephants of the Congo basin, often assumed to
be a separate species, Loxodonta pumilio Noack 1906,
are probably forest elephants whose diminutive size and/
or early maturity is due to environmental conditions
(Debruyne et al. 2003). Nevertheless, Basilio (1962) gave
an accurate description of the latter supposedly existing
form, which he considered to be widespread in the plu-
vial forest of west-central Africa, from the Congo and
Gabon to Guinea and southern Cameroon (Figure 10).
Not long ago, researchers of the authority of Martin
Eisentraut and Wolfgang Bo¨ hme (Eisentraut and Bo¨ hme
1989, Bo¨ hme and Eisentraut 1990) have also come out
in favour of the existence of this pygmy form, still dif-
fused in the Congo basin and possibly identifiable as L.
pumilio. Addressing the principal objections against the
specific status of the African pygmy elephants (e.g., eco-
logical ‘‘race’’, ecotype, juvenile of L. cyclotis), they sys-
tematically refuted them on the basis of morphological,
craniological and biological evidence. As for the size of
these pygmy proboscideans, for the first time they
offered photographic documentation – obtained in the
north of the People’s Republic of Congo – of the objec-
tive scale by juxtaposing forest elephants with dwarf buf-
faloes and above all with a great white egret, Egretta alba
L. 1758 (Bo¨ hme and Eisentraut 1990) (Figure 11).
Predators in unbalanced faunas
A peculiar aspect of the endemic insular mammalian fau-
na is that they were all unbalanced. In fact, entire taxo-
nomic groups, such as perissodactyls and felids, are
almost completely absent (Masseti and Mazza 1996).
Outside the Mediterranean, islands with unbalanced
Pleistocene faunas occur in the Japanese archipelago,
Indonesia, the Canary islands, Madagascar and off the
coast of California. However, the Mediterranean island
faunas have been defined as unbalanced mainly due to
the general absence of carnivores of large body size
Figure 10 Adult male pygmy elephant killed in the pluvial forest form Genetta plesictoides, intermediate between the
of Spanish Guinea (extant Equatorial Guinea) in September 1957
(from Basilio 1962).
Figure 11 Sketch reconstructing the size of an adult Congo
forest pygmy elephant and a calf, compared with the dimensions
of an individual of white egret, Egretta alba L. 1758 (from Bo¨ hme
and Eisentraut 1990).
(Sondaar 1977, Dermitzakis and Sondaar 1978). Willem-
sen (1992), followed by Theodorou et al. (2007a), tries to
explain this feature of the faunal composition by observ-
ing that carnivores are worse swimmers than the other
taxa that dominated the endemic faunas. Schu¨ le (1993)
also observed that islands of less than 20,000 km2 do
not support carnivores larger than an ermine, Mustela
erminea L. 1758. Furthermore, Raia and Meiri (2006)
observed that competition and predation have little or no
effect on insular carnivore body size, which is influenced
by the nature of the resource base, dependent on the
abundance and size of their prey.
However, it can be generally stated that all the carni-
vores of the unbalanced insular faunas show strong
endemic patterns and almost all of them are character-
istic of a different insular complex. Indigenous to Sardinia
and Corsica (Malatesta 1962, 1970, Bonifay 1976), Cyno-
terium sardous Studiati 1857, for example, is the only
endemic canid of the Late Pleistocene known for the
Mediterranean islands. It was approximately the size of
a fox, but its origin is still uncertain (cf. van der Geer
1988, Eisemann 1990, Lyras et al. 2006). According to
Eisemann and van der Geer (1999), craniological com-
parisons do not indicate a close relationship with any of
the extant canids, supporting a full generic rank for this
species. The occurrence of C. sardous in the Sardo-Cor-
sican insular complex may have been made possible by
the availability of prey of a size and quantity sufficient to
offer a permanent food supply. Here, in fact, the pres-
ence of the ocothonid rat-like hare, Prolagus sardus
Wagner 1829 (Figure 12), an endemic sort of pika, fairly
large in size, with a high reproduction rate, and not avail-
able on any of the other contemporary Mediterranean
islands, fulfilled the essential requirement for a medium-
sized terrestrial predator (cf. Sondaar et al. 1986).
On the other side of the Mediterranean basin, in
Cyprus, Bate (1903b) described a fossil mandibular
ramus of a Pleistocene island carnivore as the endemic
extant viverrid G. genetta (L. 1758), and a mustelid of the
European Oligocene, Plesictis croizeti Pomel 1846. Little
is known of G. plesictoides, with scant material solely
from a few Cypriot localities. Recent excavations in the