metapopulation structure might bestow better
prospects of survival. The captive flock, of about 58
birds, is maintained within the same forest area as
the wild population. In this case, the suggestion was
that on subdividing the captive flock, one of these
populations should be transferred to a multispecies
facility in the continental United States, with indi-
viduals subsequently exchanged between the popu-
lations. Wilson and colleagues pointed out the
danger of the accidental spread of disease, picked
up in this multispecies context, back to Puerto Rico.
They also pointed out that mate selection is idiosyn-
cratic in this, as in some other K-selected species,
and that subdividing the population while it was
still at a small size might be counter-productive for
this reason also.
As with the other models considered in this
chapter, it is important to understand that meta-
population models represent simplified abst-
ractions, allowing ‘what-if’ scenarios to be explored
(Harrison 1994). It is therefore dangerous to assign
primacy to any single model in determining con-
servation policy.
The core–sink model variant
The classic metapopulation models assume equiva-
lence in size and quality of patches, with all patches
susceptible to extinction and all occupied patches
able to act as re-supply centres (Fig. 10.4). Dawson
(1994) noted two common scenarios that are inade-
quately dealt with in such simplified models. First,
if one patch is big, and the others small, as at Jasper
Ridge, then they may have a core–satelliteor
source–sink relationship, in which the core is
effectively immortal and the smaller patches are
essentially sinks (Table 10.1). In a study of orb
spiders on Bahamian islands, larger populations
were found to persist, whereas small populations
repeatedly went extinct and re-immigrated from
the larger areas and then become extinct again
(Schoener and Spiller 1987), thus contributing little
or nothing to the chances of persistence of the
whole metapopulation (see also Kindvall and
Ahlen 1992; Harrison 1994). Dawson (1994)
concluded that it is usually survival on the large
habitat fragments that ensures species survival, not
an equilibrium turnover on many smaller patches
(seeThomaset al. 1996 for a counter-example).
Secondly,even seemingly isolated populations may
be saved from extinction through continued sup-
plementation by individuals migrating from the
core population. In other words, the ‘rescue effect’
may operate. Gotelli (1991) has shown that building
this effect into metapopulation scenarios has a
highly significant impact on the outcome.METAPOPULATION DYNAMICS 261(a) Classic metapopulation(b) Source-sinkFigure 10.4Metapopulation scenarios. (a) In classic models,
habitat patches are viewed as similar in size and nature: the
occupied patches (hatched) will re-supply patches that have lost
their populations (unshaded), as indicated by the arrows, and at a
later point in time will, in turn, lose their population and be
re-supplied. (b) Much more commonly, large, effectively immortal
patches (source or ‘mainland’ habitat islands) (dense shading)
re-supply smaller satellite or sink-habitat patches after density-
independent population crashes (e.g. related to adverse weather):
close and larger patches may be anticipated to be swiftly re-occupied
(hatched) but more distant patches take longer to be re-occupied,
especially if small (unshaded).