Conser vation biologists often must make management decisions based on little empirical information. In Germany, biologists are concerned that the recover y and reintroduction of Eurasian lynx (Lynx lynx) may fail because the remaining suitable habitat may be insufficient to sustain a viable population. However, no comprehensive study ad- dressing this concern has been made that not only considers distribution of suitable habitat, but also connectivity to other populations. The aims of this study were (1) to quantify the amount and location of potentially suitable lynx habitat in Germany, (2) to estimate the connectivity between patches of suitable habitat, and (3) to evaluate lynx conser vation programs. Habitat preferences of lynx were described in a rule-based model based on the availability of forest cover (defined by patch size) and the spatial structure of the habitat. Rules were implemented in a geographic information system to predict locations of suitable habitat. Optimal connections among patches were modeled using a cost-path analysis based on habitat-specific probabilities of lynx crossing patches. Results indicated wide variation in the size of patches of suitable habitat, with 10 areas each sufficiently large to sustain >20 resident lynxes. Overall, a total of 380 lynxes could be sustained by the 10 areas. Uncertainty analyses of model parameters and assumptions revealed little variation in pre- dicted habitat, primarily because results were constrained by the actual distribution of forest habitat. Our analyses suggest that lynx reintroduction programs should emphasize large, connected areas and consider broad-scale habitat connectivity in the landscape. Our ap- proach also demonstrates how biologically plausible rules can be applied in conser vation to identify areas in which success is most likely, even when few empirical data are available

Peer reviewed