G. Jansson et P. Angelstam, Threshold levels of habitat composition for the presence of the long-tailed tit (Aegithalos caudatus) in a boreal landscape, LANDSC ECOL, 14(3), 1999, pp. 283-290
We assessed the habitat patch occupancy of a deciduous-mixed forest special
ist, the long-tailed tit (Aegithalos caudatus), in a 1000 km(2) conifer dom
inated landscape in relation to two landscape parameters, namely proportion
and isolation of suitable habitat. Data from five consecutive spring seaso
ns were used and within habitat variation controlled for. The occurrence of
long-tailed tits was positively related to the amount of habitat within 1
km(2) (p=0.0007) and negatively related to the distance between habitat pat
ches (p < 0.0001). When combined, the two variables explained > 78% of the
variation in local patch occupancy. There were distinct thresholds in these
landscape variables for the probability of local long-tailed tit presence.
In the model the probability increased from 0.1 to 0.8 when interpatch dis
tance decreased from 500 to 100 m with 5% total habitat coverage. With a to
tal proportion of 15% suitable habitat, the same probability jump occurred
when interpatch distance changed from 900 to 500 m. The general importance
of defined measurements and quantified threshold levels for species conserv
ation and landscape management is discussed.