1. In a previous paper (Goss-Custard et al. 1995), a demographic model
suggested that different breeding subpopulations of a migrating shore
bird, the oystercatcher Haematopus ostralegus, that shared the same wi
ntering area might be affected in different ways by the progressive re
duction of their winter habitat. Initially, as winter habitat was remo
ved in simulations, all subpopulations decreased in parallel. However,
as habitat loss continued, the subpopulations with lower fledgling pr
oduction began to be disproportionately affected. This paper explores
what caused this difference in response between subpopulations. 2. A s
implified version of the model was used in which two or four subpopula
tions wintered in one area. Two simple density-dependent functions wer
e included, one for breeding territoriality and one for winter mortali
ty, with density dependence starting only above a certain threshold de
nsity. Each pair that acquired a breeding territory reared a constant
number of hedged young each year. 3. As winter habitat was reduced, th
e reduction in population size of the subpopulations with lower reprod
uctive rates began to accelerate as soon as the point was reached when
all adults were breeding. Before this point, breeding birds that had
died in the winter had been replaced by non-breeders. After this point
, the subpopulation reproductive output was not sufficient to replace
birds dying in the winter. At the same time, winter mortality rates re
mained high as winter densities were sustained by birds from the other
subpopulations. 4. The results suggest that, in order to detect the f
irst effects of winter habitat loss, conservationists should pay speci
al attention to monitoring breeding numbers in local subpopulations wi
th below average rates of reproduction. The results also suggest that
the cause of a decline in some local breeding subpopulations, when oth
ers remain stable, may need to be sought on the wintering grounds and
not in the breeding localities themselves.