by feral cats. All known kakapo (i.e. the entire
species) were therefore transferred to three preda-
tor-free island refuges, where, supported by sup-
plementary feeding, the kakapo appeared to have
better prospects. However, recovery was reported
to have been initially disappointing due to the bias
of survivors to older birds, a delay in resuming
breeding after translocation, and periodic invasions
of stoats on one of the islands.
Assessing the success of 45 release programmes,
Lovegrove (1996) concluded that failures occurred
where either (1) predators were still present, or
arrived after release, or (2) too few birds, with an
unbalanced sex ratio, were released. All releases
involving more than 15 birds on islands lacking
predators were successful, at least in the short term.
Other New Zealand studies reportedly show that
releases of as few as 5 birds on predator-free islands
are likely to succeed, and, more generally, the
review of Griffith et al. (1989) indicates that releases
of 40 or more birds into good habitat are generally
successful.
Clearly, the elimination of exotic predatory mam-
mals is essential. Moreover, as on occasion even
individual animals can deal a heavy blow to small
populations of threatened species, it may be neces-
sary to monitor islands closely to ensure that pred-
ators do not return. Much effort has been
channelled into developing effective methods of
predator control. Conservationists have been devel-
oping a wide range of techniques, often at consid-
erable expense, including the introduction of new
rodenticides, and the spreading of bait from heli-
copters, in order to target particular pest species to
greatest effect (e.g. Clout and Craig 1995; Micol and
Jouventin 1995). Such schemes are often unafford-
able (Herrero 1997).
It is obvious that it is important to translocate
populations into suitable habitat, but what this may
be is not so obvious, as for many species popula-
tions the reasons for their original demise are not
fully understood (e.g. Hambler 1994). Given the
role of humans in the demise of island species, it is
important that translocations take place into areas
in which harmful human influences are appropri-
ately managed, or mitigated. Because the prehis-
toric human impact was so intense in Polynesia,
Franklin and Steadman (1991) argue that the lack of
many taxa from uninhabited, forested islands may
be due to events that occurred centuries ago, and
that with human abandonment and forest regener-
ation, such islands may once again be able to sup-
port the extirpated species. They therefore advocate
the translocation of species into such islands, fol-
lowing careful assessment of the resources and
problems involved. From their preliminary assess-
ments they argue that low makatea islands, such as
the Cooks, hold limited potential, but that there
may be greater potential among the 20 or so Tongan
islands that are both uninhabited and greater than
2km^2 in area. Moreover, the fossil record demon-
strates the former presence on one such island,
‘Eua, of species that still survive in Fiji, Samoa, and
other Tongan islands, including the parrots Vini
australisandPhigys solitarius. Establishing multi-
ple populations will greatly improve the prospects
of a species surviving events such as the loss of a
local population in a hurricane. Such initiatives
require a sound biogeographical and ecological
baseline, together with the political and legal
frameworks to support the efforts.
The only wild giant tortoises in the Indian Ocean,
Geochelone gigantea, survive on Aldabra atoll. Here,
the culling of goats has enabled their population to
persist in reasonable numbers. Between 1978 and
1982, 250 tortoises were translocated to Curieuse in
the granitic Seychelles (Hambler 1994), but in a
detailed survey in 1990 it was found that only 117
animals remained. Poaching probably accounted
for most of the losses, although some of the surviv-
ing adults were diseased and low growth and
reproductive rates suggested possible saturation of
resources and nutrient limitation (particularly cal-
cium deficiencies). Both feral cats and feral rats are
abundant on the island, and probably account for
the very low rate of recruitment. To ensure the per-
sistence of the new colony, it became necessary to
establish a facility to rear hatchlings until they are
over 5 years old and thus relatively safe from pre-
dation (if not from poaching). In contrast, a translo-
cation to another of the granitic Seychelles, Frégate,
has been comparatively successful. The reasons for
the difference in outcome on the two islands are as
yet uncertain, but it is likely that both habitat332 ISLAND REMEDIES: THE CONSERVATION OF ISLAND ECOSYSTEMS