Biology Reference
In-Depth Information
migration between patches but most individuals interacting with individ-
uals in the same patch; (2) single populations are too small to have a long
lifetime relative to that of the metapopulation; (3) the patches are close
enough to permit recolonization; and (4) local dynamics are insufficiently
synchronized for a likely occurrence of simultaneous extinction of all
local populations (see also Hanski and Gilpin
1991
).
Larvae of the species have two host plants in the area. Females lay eggs
in groups of up to 200. In 1993-97 there were 300-500 local populations,
some very small but most with 50-100 larvae in late summer; some
30-50% of larval groups in the larger populations were estimated to
have not yet been discovered. Habitat patches containing the local
populations are well delimited by the surrounding environment.
Estimates showed that about 80% of the butterflies spent their entire life
in the natal patch. Even the largest population of about 650 butterflies
became extinct after a few years. Mark-recapture experiments showed
that about 9% of 741 recaptures were from a new patch. About 15% of
males and 30% of females moved to another patch during their lifetime.
Patch area, density of flowers and open patch boundaries determined
emigration and immigration rates. In a study of 1737 marked butterflies,
most migrating butterflies moved only a few hundred (less than 500) m,
although up to 3.1 km migration distance was recorded. Among 906
colonizations observed between 1994 and 2000, the mean distance from
the nearest population was 0.6 km and the longest 6.8 km (van Nouhuys
and Hanski
2002
). The annual colonization rate of empty patches was
approximately 15%.
An intensive study from 1991 to 1993 demonstrated largely asynchron-
ous local dynamics. Asynchrony appears to be maintained by an inter-
action of weather and habitat quality, and the actions of two
hymenopteran parasitoids, one of which also has distinct metapopulation
structure. One of the parasitoids (which have several hyperparasitoids)
caused larval mortality from 0-100%, varying between populations and
from year to year, indicating a very great spatial and temporal variance.
This means that despite possible density- dependent regulatory mechanisms,
the mostly very small local populations and even the largest populations,
have a high risk of extinction. Thus, between 1993 and 1995 more than
400 population extinctions were recorded. There are many suitable but
empty habitats (meadows). Application of the incidence function
model (IFM) showed that fluctuations in the proportion of occupied
patches over time were the result of spatially uncorrelated stochasticity
in extinctions and colonizations (although in real systems environmental
