Since 1990 we have initiated a number of studies on the population eco
logy, population genetics and social organization of several animal sp
ecies in forest fragments of varying sizes. Study species include squi
rrels, small passerines and butterflies. Study sites are forest fragme
nts between 1 and 40 ha, and study plots in ''continuous'' forest of 1
00 to 1500 ha. Our results so far suggest that forest fragmentation ha
s more profound effects on dispersal patterns and population structure
than on reproductive output and survival. Effects on dispersal includ
e at least two aspects, which are not necessarily found in the same sp
ecies. First, immigration deficits may be caused by low disperser succ
ess and/or source-sink effects. Second, reduced gene flow among local
populations may result in genetic differentiation and/or loss of genet
ic variation. Negative consequences of the latter are not obvious, but
indirect effects may be important e.g. for local adaptation. We concl
ude that studies on population structure in fragmented habitats should
include quantitative (dispersal distance, immi/emigration, genetic st
ructure) as well as qualitative aspects (timing of dispersal, habitat
selection).