292 ANTHROPOGENIC LOSSES AND THREATS TO ISLAND ECOSYSTEMS
If this involves very tight mutualisms (e.g. as can
be involved in pollination services), then the sur-
viving co-evolutionary partner, or ‘widow’ species
(sensuOlesen and Valido 2004) may in time follow
the same fate, unless it is able to find similar serv-
ices from another locally available species. Such
losses may then multiply, as other mutualisms are
interrupted, in a fashion analogous to the meso-
predator release mechanism discussed in the pre-
vious chapter. Such trophic cascades may follow
from both deterministic and stochastic extinctions.
This term implies a cascading effect across trophic
boundaries as an ecosystem unravels. There are
some spectacular examples driven by ‘ecosystem
transformer’ species (e.g. Rodda et al. 1999;
O’Dowdet al. 2003), but it is by no means clear
how often and how far such cascading effects oper-
ate. Many ecosystems in practice have a degree of
ecological redundancy (Olesen and Jordano 2002),
such that a cascading ‘ecological meltdown’ may
be relatively uncommon.
11.3 The scale of island losses globally
Species populations go extinct locally all the time.
Of itself, a population extinction event is not neces-
sarily of conservation concern. It becomes a matter
of concern when it marks a pattern of range col-
lapse, and especially when it represents the loss of
an endemic subspecies or species. In discussing
island species losses, we can distinguish different
levels of extinction by the terms local species extinc-
tion or island extirpation, as distinct from the final
extinction of the last population of a species, i.e.
when an island endemic form is entirely eliminated.
We focus on the latter unless otherwise denoted.
Those collating records of species extinctions
commonly take the historic period of island explo-
ration and scientific record to have begun c.AD1600.
In the period so defined, significantly more species
of plants and animals are known to have become
extinct from islands than from continents: for the
taxa listed in Table 11.2, about 60% have been island
species. However, we know so little about the
losses of invertebrates that it would probably be
safer to discount them from the analysis. For the
groups with the best data, the proportion of extinc-
tions from islands varies from 60% for mammals
(which are generally lacking from remote islands in
the first place), to 79% for molluscs, 81% for birds,
and as high as 95% for reptiles. Moreover, of 121
continental species losses cited by Groombridge
(1992), about 66% can be classified as aquatic
species, many of them inhabiting lakes, which can
be regarded as ‘negative islands’ (Box 11.1).
The values in Table 11.2 reflect a highly cautious
approach to counting extinctions, and these num-
bers are predicted to change radically over the com-
ing decades as the various ‘extinction debts’
(Chapter 10) are called in the world over. To date,
therefore, relatively small numbers of recent terres-
trial animal extinctions have been recorded from
mainland tropical forest regions such as Amazonia
and the Atlantic forests of Brazil, although of course
these are the areas from which large numbers of
extinctions are estimated to be occurring or about to
occur. Yet, when we place the figures for island
extinctions and threat into the context of how many
species originate from islands, it is clear that they
have suffered disproportionate losses in recent time
and also that they remain priority sites for conserva-
tion. In the case of birds, and allowing for theTable 11.1Largely stochastic, natural causes of extinction, versus deterministic causes arising from human activities on islands
Stochastic causes (inherent to islands) Deterministic causes (related to human activities)
Natural disasters (volcanic eruptions, landslides, hurricanes, Habitat loss, degradation, or fragmentation
tsunamis)
Taxon cycle dynamics (including demographic collapse Alien species introduction (including competitors, predators,
and inbreeding depression) hybridizing congeners, parasites, disease vectors, or diseases)
Pleistocene climatic fluctuations (including sea-level rise) Direct predation (including hunting, fishing, or specimen collection)
—— Trophic cascade ——